Add a check ensuring there is no modification in the fid dbs

This reverts commit b12ffd1356.

.github/workflows/check-valid-milestone.yml

@@ -0,0 +1,100 @@
+name: PR Milestone Check
+
+on:
+  pull_request:
+    types: [opened, reopened, edited, synchronize, milestoned, demilestoned]
+    branches:
+      - "main"
+      - "release-v*.*.*"
+
+jobs:
+  check-milestone:
+    name: Check PR Milestone
+    runs-on: ubuntu-latest
+
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v3
+
+      - name: Validate PR milestone
+        uses: actions/github-script@v6
+        with:
+          github-token: ${{ secrets.GITHUB_TOKEN }}
+          script: |
+            // Get PR number directly from the event payload
+            const prNumber = context.payload.pull_request.number;
+
+            // Get PR details
+            const { data: prData } = await github.rest.pulls.get({
+              owner: 'meilisearch',
+              repo: 'meilisearch',
+              pull_number: prNumber
+            });
+
+            // Get base branch name
+            const baseBranch = prData.base.ref;
+            console.log(`Base branch: ${baseBranch}`);
+
+            // Get PR milestone
+            const prMilestone = prData.milestone;
+            if (!prMilestone) {
+              core.setFailed('PR must have a milestone assigned');
+              return;
+            }
+            console.log(`PR milestone: ${prMilestone.title}`);
+
+            // Validate milestone format: vx.y.z
+            const milestoneRegex = /^v\d+\.\d+\.\d+$/;
+            if (!milestoneRegex.test(prMilestone.title)) {
+              core.setFailed(`Milestone "${prMilestone.title}" does not follow the required format vx.y.z`);
+              return;
+            }
+
+            // For main branch PRs, check if the milestone is the highest one
+            if (baseBranch === 'main') {
+              // Get all milestones
+              const { data: milestones } = await github.rest.issues.listMilestones({
+                owner: 'meilisearch',
+                repo: 'meilisearch',
+                state: 'open',
+                sort: 'due_on',
+                direction: 'desc'
+              });
+
+              // Sort milestones by version number (vx.y.z)
+              const sortedMilestones = milestones
+                .filter(m => milestoneRegex.test(m.title))
+                .sort((a, b) => {
+                  const versionA = a.title.substring(1).split('.').map(Number);
+                  const versionB = b.title.substring(1).split('.').map(Number);
+
+                  // Compare major version
+                  if (versionA[0] !== versionB[0]) return versionB[0] - versionA[0];
+                  // Compare minor version
+                  if (versionA[1] !== versionB[1]) return versionB[1] - versionA[1];
+                  // Compare patch version
+                  return versionB[2] - versionA[2];
+                });
+
+              if (sortedMilestones.length === 0) {
+                core.setFailed('No valid milestones found in the repository. Please create at least one milestone with the format vx.y.z');
+                return;
+              }
+
+              const highestMilestone = sortedMilestones[0];
+              console.log(`Highest milestone: ${highestMilestone.title}`);
+
+              if (prMilestone.title !== highestMilestone.title) {
+                core.setFailed(`PRs targeting the main branch must use the highest milestone (${highestMilestone.title}), but this PR uses ${prMilestone.title}`);
+                return;
+              }
+            } else {
+              // For release branches, the milestone should match the branch version
+              const branchVersion = baseBranch.substring(8); // remove 'release-'
+              if (prMilestone.title !== branchVersion) {
+                core.setFailed(`PRs targeting release branch "${baseBranch}" must use the matching milestone "${branchVersion}", but this PR uses "${prMilestone.title}"`);
+                return;
+              }
+            }
+
+            console.log('PR milestone validation passed!');

.github/workflows/flaky-tests.yml

@@ -9,22 +9,22 @@ jobs:
   flaky:
     runs-on: ubuntu-latest
     container:
-      # Use ubuntu-20.04 to compile with glibc 2.28
-      image: ubuntu:20.04
+      # Use ubuntu-22.04 to compile with glibc 2.35
+      image: ubuntu:22.04
     steps:
-    - uses: actions/checkout@v3
-    - name: Install needed dependencies
-      run: |
-        apt-get update && apt-get install -y curl
-        apt-get install build-essential -y
-    - uses: dtolnay/rust-toolchain@1.81
-    - name: Install cargo-flaky
-      run: cargo install cargo-flaky
-    - name: Run cargo flaky in the dumps
-      run: cd crates/dump; cargo flaky -i 100 --release
-    - name: Run cargo flaky in the index-scheduler
-      run: cd crates/index-scheduler; cargo flaky -i 100 --release
-    - name: Run cargo flaky in the auth
-      run: cd crates/meilisearch-auth; cargo flaky -i 100 --release
-    - name: Run cargo flaky in meilisearch
-      run: cd crates/meilisearch; cargo flaky -i 100 --release
+      - uses: actions/checkout@v3
+      - name: Install needed dependencies
+        run: |
+          apt-get update && apt-get install -y curl
+          apt-get install build-essential -y
+      - uses: dtolnay/rust-toolchain@1.81
+      - name: Install cargo-flaky
+        run: cargo install cargo-flaky
+      - name: Run cargo flaky in the dumps
+        run: cd crates/dump; cargo flaky -i 100 --release
+      - name: Run cargo flaky in the index-scheduler
+        run: cd crates/index-scheduler; cargo flaky -i 100 --release
+      - name: Run cargo flaky in the auth
+        run: cd crates/meilisearch-auth; cargo flaky -i 100 --release
+      - name: Run cargo flaky in meilisearch
+        run: cd crates/meilisearch; cargo flaky -i 100 --release

.github/workflows/publish-apt-brew-pkg.yml

@@ -18,28 +18,28 @@ jobs:
     runs-on: ubuntu-latest
     needs: check-version
     container:
-      # Use ubuntu-20.04 to compile with glibc 2.28
-      image: ubuntu:20.04
+      # Use ubuntu-22.04 to compile with glibc 2.35
+      image: ubuntu:22.04
     steps:
-    - name: Install needed dependencies
-      run: |
-        apt-get update && apt-get install -y curl
-        apt-get install build-essential -y
-    - uses: dtolnay/rust-toolchain@1.81
-    - name: Install cargo-deb
-      run: cargo install cargo-deb
-    - uses: actions/checkout@v3
-    - name: Build deb package
-      run: cargo deb -p meilisearch -o target/debian/meilisearch.deb
-    - name: Upload debian pkg to release
-      uses: svenstaro/upload-release-action@2.7.0
-      with:
-        repo_token: ${{ secrets.MEILI_BOT_GH_PAT }}
-        file: target/debian/meilisearch.deb
-        asset_name: meilisearch.deb
-        tag: ${{ github.ref }}
-    - name: Upload debian pkg to apt repository
-      run: curl -F package=@target/debian/meilisearch.deb https://${{ secrets.GEMFURY_PUSH_TOKEN }}@push.fury.io/meilisearch/
+      - name: Install needed dependencies
+        run: |
+          apt-get update && apt-get install -y curl
+          apt-get install build-essential -y
+      - uses: dtolnay/rust-toolchain@1.81
+      - name: Install cargo-deb
+        run: cargo install cargo-deb
+      - uses: actions/checkout@v3
+      - name: Build deb package
+        run: cargo deb -p meilisearch -o target/debian/meilisearch.deb
+      - name: Upload debian pkg to release
+        uses: svenstaro/upload-release-action@2.7.0
+        with:
+          repo_token: ${{ secrets.MEILI_BOT_GH_PAT }}
+          file: target/debian/meilisearch.deb
+          asset_name: meilisearch.deb
+          tag: ${{ github.ref }}
+      - name: Upload debian pkg to apt repository
+        run: curl -F package=@target/debian/meilisearch.deb https://${{ secrets.GEMFURY_PUSH_TOKEN }}@push.fury.io/meilisearch/
 
   homebrew:
     name: Bump Homebrew formula

.github/workflows/publish-binaries.yml

@@ -3,7 +3,7 @@ name: Publish binaries to GitHub release
 on:
   workflow_dispatch:
   schedule:
-    - cron: '0 2 * * *' # Every day at 2:00am
+    - cron: "0 2 * * *" # Every day at 2:00am
   release:
     types: [published]
 
@@ -37,26 +37,26 @@ jobs:
     runs-on: ubuntu-latest
     needs: check-version
     container:
-      # Use ubuntu-20.04 to compile with glibc 2.28
-      image: ubuntu:20.04
+      # Use ubuntu-22.04 to compile with glibc 2.35
+      image: ubuntu:22.04
     steps:
-    - uses: actions/checkout@v3
-    - name: Install needed dependencies
-      run: |
-        apt-get update && apt-get install -y curl
-        apt-get install build-essential -y
-    - uses: dtolnay/rust-toolchain@1.81
-    - name: Build
-      run: cargo build --release --locked
-    # No need to upload binaries for dry run (cron)
-    - name: Upload binaries to release
-      if: github.event_name == 'release'
-      uses: svenstaro/upload-release-action@2.7.0
-      with:
-        repo_token: ${{ secrets.MEILI_BOT_GH_PAT }}
-        file: target/release/meilisearch
-        asset_name: meilisearch-linux-amd64
-        tag: ${{ github.ref }}
+      - uses: actions/checkout@v3
+      - name: Install needed dependencies
+        run: |
+          apt-get update && apt-get install -y curl
+          apt-get install build-essential -y
+      - uses: dtolnay/rust-toolchain@1.81
+      - name: Build
+        run: cargo build --release --locked
+      # No need to upload binaries for dry run (cron)
+      - name: Upload binaries to release
+        if: github.event_name == 'release'
+        uses: svenstaro/upload-release-action@2.7.0
+        with:
+          repo_token: ${{ secrets.MEILI_BOT_GH_PAT }}
+          file: target/release/meilisearch
+          asset_name: meilisearch-linux-amd64
+          tag: ${{ github.ref }}
 
   publish-macos-windows:
     name: Publish binary for ${{ matrix.os }}
@@ -74,19 +74,19 @@ jobs:
             artifact_name: meilisearch.exe
             asset_name: meilisearch-windows-amd64.exe
     steps:
-    - uses: actions/checkout@v3
-    - uses: dtolnay/rust-toolchain@1.81
-    - name: Build
-      run: cargo build --release --locked
-    # No need to upload binaries for dry run (cron)
-    - name: Upload binaries to release
-      if: github.event_name == 'release'
-      uses: svenstaro/upload-release-action@2.7.0
-      with:
-        repo_token: ${{ secrets.MEILI_BOT_GH_PAT }}
-        file: target/release/${{ matrix.artifact_name }}
-        asset_name: ${{ matrix.asset_name }}
-        tag: ${{ github.ref }}
+      - uses: actions/checkout@v3
+      - uses: dtolnay/rust-toolchain@1.81
+      - name: Build
+        run: cargo build --release --locked
+      # No need to upload binaries for dry run (cron)
+      - name: Upload binaries to release
+        if: github.event_name == 'release'
+        uses: svenstaro/upload-release-action@2.7.0
+        with:
+          repo_token: ${{ secrets.MEILI_BOT_GH_PAT }}
+          file: target/release/${{ matrix.artifact_name }}
+          asset_name: ${{ matrix.asset_name }}
+          tag: ${{ github.ref }}
 
   publish-macos-apple-silicon:
     name: Publish binary for macOS silicon
@@ -127,8 +127,8 @@ jobs:
     env:
       DEBIAN_FRONTEND: noninteractive
     container:
-      # Use ubuntu-20.04 to compile with glibc 2.28
-      image: ubuntu:20.04
+      # Use ubuntu-22.04 to compile with glibc 2.35
+      image: ubuntu:22.04
     strategy:
       matrix:
         include:

.github/workflows/sdks-tests.yml

@@ -52,7 +52,7 @@ jobs:
       - name: Setup .NET Core
         uses: actions/setup-dotnet@v4
         with:
-          dotnet-version: "6.0.x"
+          dotnet-version: "8.0.x"
       - name: Install dependencies
         run: dotnet restore
       - name: Build

.github/workflows/test-suite.yml

@@ -19,13 +19,13 @@ env:
 
 jobs:
   test-linux:
-    name: Tests on ubuntu-20.04
+    name: Tests on ubuntu-22.04
     runs-on: ubuntu-latest
     container:
-      # Use ubuntu-20.04 to compile with glibc 2.28
-      image: ubuntu:20.04
+      # Use ubuntu-22.04 to compile with glibc 2.35
+      image: ubuntu:22.04
     steps:
-      - uses: actions/checkout@v3
+      - uses: actions/checkout@v4
       - name: Install needed dependencies
         run: |
           apt-get update && apt-get install -y curl
@@ -72,8 +72,8 @@ jobs:
     name: Tests almost all features
     runs-on: ubuntu-latest
     container:
-      # Use ubuntu-20.04 to compile with glibc 2.28
-      image: ubuntu:20.04
+      # Use ubuntu-22.04 to compile with glibc 2.35
+      image: ubuntu:22.04
     if: github.event_name == 'schedule' || github.event_name == 'workflow_dispatch'
     steps:
       - uses: actions/checkout@v3
@@ -125,7 +125,7 @@ jobs:
     name: Test disabled tokenization
     runs-on: ubuntu-latest
     container:
-      image: ubuntu:20.04
+      image: ubuntu:22.04
     if: github.event_name == 'schedule' || github.event_name == 'workflow_dispatch'
     steps:
       - uses: actions/checkout@v3
@@ -149,8 +149,8 @@ jobs:
     name: Run tests in debug
     runs-on: ubuntu-latest
     container:
-      # Use ubuntu-20.04 to compile with glibc 2.28
-      image: ubuntu:20.04
+      # Use ubuntu-22.04 to compile with glibc 2.35
+      image: ubuntu:22.04
     steps:
       - uses: actions/checkout@v3
       - name: Install needed dependencies

CONTRIBUTING.md

@@ -95,6 +95,11 @@ Meilisearch follows the [cargo xtask](https://github.com/matklad/cargo-xtask) wo
 
 Run `cargo xtask --help` from the root of the repository to find out what is available.
 
+#### Update the openAPI file if the API changed
+
+To update the openAPI file in the code, see [sprint_issue.md](https://github.com/meilisearch/meilisearch/blob/main/.github/ISSUE_TEMPLATE/sprint_issue.md#reminders-when-modifying-the-api).
+If you want to update the openAPI file on the [open-api repository](https://github.com/meilisearch/open-api), see [update-openapi-issue.md](https://github.com/meilisearch/engine-team/blob/main/issue-templates/update-openapi-issue.md).
+
 ### Logging
 
 Meilisearch uses [`tracing`](https://lib.rs/crates/tracing) for logging purposes. Tracing logs are structured and can be displayed as JSON to the end user, so prefer passing arguments as fields rather than interpolating them in the message.

Cargo.lock

@@ -503,7 +503,7 @@ source = "git+https://github.com/meilisearch/bbqueue#cbb87cc707b5af415ef203bdaf2
 
 [[package]]
 name = "benchmarks"
-version = "1.13.0"
+version = "1.13.3"
 dependencies = [
  "anyhow",
  "bumpalo",
@@ -694,7 +694,7 @@ dependencies = [
 
 [[package]]
 name = "build-info"
-version = "1.13.0"
+version = "1.13.3"
 dependencies = [
  "anyhow",
  "time",
@@ -1671,7 +1671,7 @@ dependencies = [
 
 [[package]]
 name = "dump"
-version = "1.13.0"
+version = "1.13.3"
 dependencies = [
  "anyhow",
  "big_s",
@@ -1873,7 +1873,7 @@ checksum = "37909eebbb50d72f9059c3b6d82c0463f2ff062c9e95845c43a6c9c0355411be"
 
 [[package]]
 name = "file-store"
-version = "1.13.0"
+version = "1.13.3"
 dependencies = [
  "tempfile",
  "thiserror 2.0.9",
@@ -1895,7 +1895,7 @@ dependencies = [
 
 [[package]]
 name = "filter-parser"
-version = "1.13.0"
+version = "1.13.3"
 dependencies = [
  "insta",
  "nom",
@@ -1915,7 +1915,7 @@ dependencies = [
 
 [[package]]
 name = "flatten-serde-json"
-version = "1.13.0"
+version = "1.13.3"
 dependencies = [
  "criterion",
  "serde_json",
@@ -2054,7 +2054,7 @@ dependencies = [
 
 [[package]]
 name = "fuzzers"
-version = "1.13.0"
+version = "1.13.3"
 dependencies = [
  "arbitrary",
  "bumpalo",
@@ -2743,7 +2743,7 @@ checksum = "206ca75c9c03ba3d4ace2460e57b189f39f43de612c2f85836e65c929701bb2d"
 
 [[package]]
 name = "index-scheduler"
-version = "1.13.0"
+version = "1.13.3"
 dependencies = [
  "anyhow",
  "arroy 0.5.0 (registry+https://github.com/rust-lang/crates.io-index)",
@@ -2950,7 +2950,7 @@ dependencies = [
 
 [[package]]
 name = "json-depth-checker"
-version = "1.13.0"
+version = "1.13.3"
 dependencies = [
  "criterion",
  "serde_json",
@@ -3569,7 +3569,7 @@ checksum = "490cc448043f947bae3cbee9c203358d62dbee0db12107a74be5c30ccfd09771"
 
 [[package]]
 name = "meili-snap"
-version = "1.13.0"
+version = "1.13.3"
 dependencies = [
  "insta",
  "md5",
@@ -3578,7 +3578,7 @@ dependencies = [
 
 [[package]]
 name = "meilisearch"
-version = "1.13.0"
+version = "1.13.3"
 dependencies = [
  "actix-cors",
  "actix-http",
@@ -3670,7 +3670,7 @@ dependencies = [
 
 [[package]]
 name = "meilisearch-auth"
-version = "1.13.0"
+version = "1.13.3"
 dependencies = [
  "base64 0.22.1",
  "enum-iterator",
@@ -3689,7 +3689,7 @@ dependencies = [
 
 [[package]]
 name = "meilisearch-types"
-version = "1.13.0"
+version = "1.13.3"
 dependencies = [
  "actix-web",
  "anyhow",
@@ -3723,7 +3723,7 @@ dependencies = [
 
 [[package]]
 name = "meilitool"
-version = "1.13.0"
+version = "1.13.3"
 dependencies = [
  "anyhow",
  "arroy 0.5.0 (git+https://github.com/meilisearch/arroy/?tag=DO-NOT-DELETE-upgrade-v04-to-v05)",
@@ -3758,7 +3758,7 @@ dependencies = [
 
 [[package]]
 name = "milli"
-version = "1.13.0"
+version = "1.13.3"
 dependencies = [
  "allocator-api2",
  "arroy 0.5.0 (registry+https://github.com/rust-lang/crates.io-index)",
@@ -4270,7 +4270,7 @@ checksum = "e3148f5046208a5d56bcfc03053e3ca6334e51da8dfb19b6cdc8b306fae3283e"
 
 [[package]]
 name = "permissive-json-pointer"
-version = "1.13.0"
+version = "1.13.3"
 dependencies = [
  "big_s",
  "serde_json",
@@ -5160,9 +5160,9 @@ dependencies = [
 
 [[package]]
 name = "serde_json"
-version = "1.0.135"
+version = "1.0.138"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "2b0d7ba2887406110130a978386c4e1befb98c674b4fba677954e4db976630d9"
+checksum = "d434192e7da787e94a6ea7e9670b26a036d0ca41e0b7efb2676dd32bae872949"
 dependencies = [
  "indexmap",
  "itoa",
@@ -6847,7 +6847,7 @@ dependencies = [
 
 [[package]]
 name = "xtask"
-version = "1.13.0"
+version = "1.13.3"
 dependencies = [
  "anyhow",
  "build-info",

Cargo.toml

@@ -22,7 +22,7 @@ members = [
 ]
 
 [workspace.package]
-version = "1.13.0"
+version = "1.13.3"
 authors = [
     "Quentin de Quelen <quentin@dequelen.me>",
     "Clément Renault <clement@meilisearch.com>",

bors.toml

@@ -1,11 +1,10 @@
 status = [
-    'Tests on ubuntu-20.04',
+    'Tests on ubuntu-22.04',
     'Tests on macos-13',
     'Tests on windows-2022',
     'Run Clippy',
     'Run Rustfmt',
     'Run tests in debug',
 ]
-pr_status = ['Milestone Check']
 # 3 hours timeout
 timeout-sec = 10800

crates/benchmarks/benches/indexing.rs

@@ -12,7 +12,7 @@ use milli::progress::Progress;
 use milli::update::new::indexer;
 use milli::update::{IndexerConfig, Settings};
 use milli::vector::EmbeddingConfigs;
-use milli::Index;
+use milli::{FilterableAttributesRule, Index};
 use rand::seq::SliceRandom;
 use rand_chacha::rand_core::SeedableRng;
 use roaring::RoaringBitmap;
@@ -57,7 +57,8 @@ fn setup_settings<'t>(
     let searchable_fields = searchable_fields.iter().map(|s| s.to_string()).collect();
     builder.set_searchable_fields(searchable_fields);
 
-    let filterable_fields = filterable_fields.iter().map(|s| s.to_string()).collect();
+    let filterable_fields =
+        filterable_fields.iter().map(|s| FilterableAttributesRule::Field(s.to_string())).collect();
     builder.set_filterable_fields(filterable_fields);
 
     let sortable_fields = sortable_fields.iter().map(|s| s.to_string()).collect();

crates/benchmarks/benches/search_geo.rs

@@ -2,7 +2,7 @@ mod datasets_paths;
 mod utils;
 
 use criterion::{criterion_group, criterion_main};
-use milli::update::Settings;
+use milli::{update::Settings, FilterableAttributesRule};
 use utils::Conf;
 
 #[cfg(not(windows))]
@@ -21,8 +21,10 @@ fn base_conf(builder: &mut Settings) {
         ["name", "alternatenames", "elevation"].iter().map(|s| s.to_string()).collect();
     builder.set_searchable_fields(searchable_fields);
 
-    let filterable_fields =
-        ["_geo", "population", "elevation"].iter().map(|s| s.to_string()).collect();
+    let filterable_fields = ["_geo", "population", "elevation"]
+        .iter()
+        .map(|s| FilterableAttributesRule::Field(s.to_string()))
+        .collect();
     builder.set_filterable_fields(filterable_fields);
 
     let sortable_fields =

crates/benchmarks/benches/search_songs.rs

@@ -2,7 +2,7 @@ mod datasets_paths;
 mod utils;
 
 use criterion::{criterion_group, criterion_main};
-use milli::update::Settings;
+use milli::{update::Settings, FilterableAttributesRule};
 use utils::Conf;
 
 #[cfg(not(windows))]
@@ -22,7 +22,7 @@ fn base_conf(builder: &mut Settings) {
 
     let faceted_fields = ["released-timestamp", "duration-float", "genre", "country", "artist"]
         .iter()
-        .map(|s| s.to_string())
+        .map(|s| FilterableAttributesRule::Field(s.to_string()))
         .collect();
     builder.set_filterable_fields(faceted_fields);
 }

crates/dump/README.md

@@ -10,8 +10,10 @@ dump
 ├── instance-uid.uuid
 ├── keys.jsonl
 ├── metadata.json
-└── tasks
-    ├── update_files
-    │   └── [task_id].jsonl
+├── tasks
+│   ├── update_files
+│   │   └── [task_id].jsonl
+│   └── queue.jsonl
+└── batches
     └── queue.jsonl
-```
+```

crates/dump/src/lib.rs

@@ -228,14 +228,16 @@ pub(crate) mod test {
 
     use big_s::S;
     use maplit::{btreemap, btreeset};
+    use meilisearch_types::batches::{Batch, BatchEnqueuedAt, BatchStats};
     use meilisearch_types::facet_values_sort::FacetValuesSort;
-    use meilisearch_types::features::RuntimeTogglableFeatures;
+    use meilisearch_types::features::{Network, Remote, RuntimeTogglableFeatures};
     use meilisearch_types::index_uid_pattern::IndexUidPattern;
     use meilisearch_types::keys::{Action, Key};
-    use meilisearch_types::milli;
     use meilisearch_types::milli::update::Setting;
+    use meilisearch_types::milli::{self, FilterableAttributesRule};
     use meilisearch_types::settings::{Checked, FacetingSettings, Settings};
-    use meilisearch_types::tasks::{Details, Status};
+    use meilisearch_types::task_view::DetailsView;
+    use meilisearch_types::tasks::{Details, Kind, Status};
     use serde_json::{json, Map, Value};
     use time::macros::datetime;
     use uuid::Uuid;
@@ -277,7 +279,10 @@ pub(crate) mod test {
         let settings = Settings {
             displayed_attributes: Setting::Set(vec![S("race"), S("name")]).into(),
             searchable_attributes: Setting::Set(vec![S("name"), S("race")]).into(),
-            filterable_attributes: Setting::Set(btreeset! { S("race"), S("age") }),
+            filterable_attributes: Setting::Set(vec![
+                FilterableAttributesRule::Field(S("race")),
+                FilterableAttributesRule::Field(S("age")),
+            ]),
             sortable_attributes: Setting::Set(btreeset! { S("age") }),
             ranking_rules: Setting::NotSet,
             stop_words: Setting::NotSet,
@@ -305,6 +310,32 @@ pub(crate) mod test {
         settings.check()
     }
 
+    pub fn create_test_batches() -> Vec<Batch> {
+        vec![Batch {
+            uid: 0,
+            details: DetailsView {
+                received_documents: Some(12),
+                indexed_documents: Some(Some(10)),
+                ..DetailsView::default()
+            },
+            progress: None,
+            stats: BatchStats {
+                total_nb_tasks: 1,
+                status: maplit::btreemap! { Status::Succeeded => 1 },
+                types: maplit::btreemap! { Kind::DocumentAdditionOrUpdate => 1 },
+                index_uids: maplit::btreemap! { "doggo".to_string() => 1 },
+                progress_trace: Default::default(),
+                write_channel_congestion: None,
+            },
+            enqueued_at: Some(BatchEnqueuedAt {
+                earliest: datetime!(2022-11-11 0:00 UTC),
+                oldest: datetime!(2022-11-11 0:00 UTC),
+            }),
+            started_at: datetime!(2022-11-20 0:00 UTC),
+            finished_at: Some(datetime!(2022-11-21 0:00 UTC)),
+        }]
+    }
+
     pub fn create_test_tasks() -> Vec<(TaskDump, Option<Vec<Document>>)> {
         vec![
             (
@@ -427,6 +458,15 @@ pub(crate) mod test {
         index.flush().unwrap();
         index.settings(&settings).unwrap();
 
+        // ========== pushing the batch queue
+        let batches = create_test_batches();
+
+        let mut batch_queue = dump.create_batches_queue().unwrap();
+        for batch in &batches {
+            batch_queue.push_batch(batch).unwrap();
+        }
+        batch_queue.flush().unwrap();
+
         // ========== pushing the task queue
         let tasks = create_test_tasks();
 
@@ -455,6 +495,10 @@ pub(crate) mod test {
 
         dump.create_experimental_features(features).unwrap();
 
+        // ========== network
+        let network = create_test_network();
+        dump.create_network(network).unwrap();
+
         // create the dump
         let mut file = tempfile::tempfile().unwrap();
         dump.persist_to(&mut file).unwrap();
@@ -467,6 +511,13 @@ pub(crate) mod test {
         RuntimeTogglableFeatures::default()
     }
 
+    fn create_test_network() -> Network {
+        Network {
+            local: Some("myself".to_string()),
+            remotes: maplit::btreemap! {"other".to_string() => Remote { url: "http://test".to_string(), search_api_key: Some("apiKey".to_string()) }},
+        }
+    }
+
     #[test]
     fn test_creating_and_read_dump() {
         let mut file = create_test_dump();
@@ -515,5 +566,9 @@ pub(crate) mod test {
         // ==== checking the features
         let expected = create_test_features();
         assert_eq!(dump.features().unwrap().unwrap(), expected);
+
+        // ==== checking the network
+        let expected = create_test_network();
+        assert_eq!(&expected, dump.network().unwrap().unwrap());
     }
 }

crates/dump/src/reader/compat/v5_to_v6.rs

@@ -196,6 +196,10 @@ impl CompatV5ToV6 {
     pub fn features(&self) -> Result<Option<v6::RuntimeTogglableFeatures>> {
         Ok(None)
     }
+
+    pub fn network(&self) -> Result<Option<&v6::Network>> {
+        Ok(None)
+    }
 }
 
 pub enum CompatIndexV5ToV6 {
@@ -318,7 +322,16 @@ impl<T> From<v5::Settings<T>> for v6::Settings<v6::Unchecked> {
         v6::Settings {
             displayed_attributes: v6::Setting::from(settings.displayed_attributes).into(),
             searchable_attributes: v6::Setting::from(settings.searchable_attributes).into(),
-            filterable_attributes: settings.filterable_attributes.into(),
+            filterable_attributes: match settings.filterable_attributes {
+                v5::settings::Setting::Set(filterable_attributes) => v6::Setting::Set(
+                    filterable_attributes
+                        .into_iter()
+                        .map(v6::FilterableAttributesRule::Field)
+                        .collect(),
+                ),
+                v5::settings::Setting::Reset => v6::Setting::Reset,
+                v5::settings::Setting::NotSet => v6::Setting::NotSet,
+            },
             sortable_attributes: settings.sortable_attributes.into(),
             ranking_rules: {
                 match settings.ranking_rules {

crates/dump/src/reader/mod.rs

@@ -23,6 +23,7 @@ mod v6;
 pub type Document = serde_json::Map<String, serde_json::Value>;
 pub type UpdateFile = dyn Iterator<Item = Result<Document>>;
 
+#[allow(clippy::large_enum_variant)]
 pub enum DumpReader {
     Current(V6Reader),
     Compat(CompatV5ToV6),
@@ -101,6 +102,13 @@ impl DumpReader {
         }
     }
 
+    pub fn batches(&mut self) -> Result<Box<dyn Iterator<Item = Result<v6::Batch>> + '_>> {
+        match self {
+            DumpReader::Current(current) => Ok(current.batches()),
+            DumpReader::Compat(_compat) => Ok(Box::new(std::iter::empty())),
+        }
+    }
+
     pub fn keys(&mut self) -> Result<Box<dyn Iterator<Item = Result<v6::Key>> + '_>> {
         match self {
             DumpReader::Current(current) => Ok(current.keys()),
@@ -114,6 +122,13 @@ impl DumpReader {
             DumpReader::Compat(compat) => compat.features(),
         }
     }
+
+    pub fn network(&self) -> Result<Option<&v6::Network>> {
+        match self {
+            DumpReader::Current(current) => Ok(current.network()),
+            DumpReader::Compat(compat) => compat.network(),
+        }
+    }
 }
 
 impl From<V6Reader> for DumpReader {
@@ -219,6 +234,10 @@ pub(crate) mod test {
         insta::assert_snapshot!(dump.date().unwrap(), @"2024-05-16 15:51:34.151044 +00:00:00");
         insta::assert_debug_snapshot!(dump.instance_uid().unwrap(), @"None");
 
+        // batches didn't exists at the time
+        let batches = dump.batches().unwrap().collect::<Result<Vec<_>>>().unwrap();
+        meili_snap::snapshot!(meili_snap::json_string!(batches), @"[]");
+
         // tasks
         let tasks = dump.tasks().unwrap().collect::<Result<Vec<_>>>().unwrap();
         let (tasks, update_files): (Vec<_>, Vec<_>) = tasks.into_iter().unzip();
@@ -328,6 +347,7 @@ pub(crate) mod test {
         }
 
         assert_eq!(dump.features().unwrap().unwrap(), RuntimeTogglableFeatures::default());
+        assert_eq!(dump.network().unwrap(), None);
     }
 
     #[test]
@@ -339,6 +359,10 @@ pub(crate) mod test {
         insta::assert_snapshot!(dump.date().unwrap(), @"2023-07-06 7:10:27.21958 +00:00:00");
         insta::assert_debug_snapshot!(dump.instance_uid().unwrap(), @"None");
 
+        // batches didn't exists at the time
+        let batches = dump.batches().unwrap().collect::<Result<Vec<_>>>().unwrap();
+        meili_snap::snapshot!(meili_snap::json_string!(batches), @"[]");
+
         // tasks
         let tasks = dump.tasks().unwrap().collect::<Result<Vec<_>>>().unwrap();
         let (tasks, update_files): (Vec<_>, Vec<_>) = tasks.into_iter().unzip();
@@ -373,6 +397,27 @@ pub(crate) mod test {
         assert_eq!(dump.features().unwrap().unwrap(), RuntimeTogglableFeatures::default());
     }
 
+    #[test]
+    fn import_dump_v6_network() {
+        let dump = File::open("tests/assets/v6-with-network.dump").unwrap();
+        let dump = DumpReader::open(dump).unwrap();
+
+        // top level infos
+        insta::assert_snapshot!(dump.date().unwrap(), @"2025-01-29 15:45:32.738676 +00:00:00");
+        insta::assert_debug_snapshot!(dump.instance_uid().unwrap(), @"None");
+
+        // network
+
+        let network = dump.network().unwrap().unwrap();
+        insta::assert_snapshot!(network.local.as_ref().unwrap(), @"ms-0");
+        insta::assert_snapshot!(network.remotes.get("ms-0").as_ref().unwrap().url, @"http://localhost:7700");
+        insta::assert_snapshot!(network.remotes.get("ms-0").as_ref().unwrap().search_api_key.is_none(), @"true");
+        insta::assert_snapshot!(network.remotes.get("ms-1").as_ref().unwrap().url, @"http://localhost:7701");
+        insta::assert_snapshot!(network.remotes.get("ms-1").as_ref().unwrap().search_api_key.is_none(), @"true");
+        insta::assert_snapshot!(network.remotes.get("ms-2").as_ref().unwrap().url, @"http://ms-5679.example.meilisearch.io");
+        insta::assert_snapshot!(network.remotes.get("ms-2").as_ref().unwrap().search_api_key.as_ref().unwrap(), @"foo");
+    }
+
     #[test]
     fn import_dump_v5() {
         let dump = File::open("tests/assets/v5.dump").unwrap();
@@ -382,6 +427,10 @@ pub(crate) mod test {
         insta::assert_snapshot!(dump.date().unwrap(), @"2022-10-04 15:55:10.344982459 +00:00:00");
         insta::assert_snapshot!(dump.instance_uid().unwrap().unwrap(), @"9e15e977-f2ae-4761-943f-1eaf75fd736d");
 
+        // batches didn't exists at the time
+        let batches = dump.batches().unwrap().collect::<Result<Vec<_>>>().unwrap();
+        meili_snap::snapshot!(meili_snap::json_string!(batches), @"[]");
+
         // tasks
         let tasks = dump.tasks().unwrap().collect::<Result<Vec<_>>>().unwrap();
         let (tasks, update_files): (Vec<_>, Vec<_>) = tasks.into_iter().unzip();
@@ -462,6 +511,10 @@ pub(crate) mod test {
         insta::assert_snapshot!(dump.date().unwrap(), @"2022-10-06 12:53:49.131989609 +00:00:00");
         insta::assert_snapshot!(dump.instance_uid().unwrap().unwrap(), @"9e15e977-f2ae-4761-943f-1eaf75fd736d");
 
+        // batches didn't exists at the time
+        let batches = dump.batches().unwrap().collect::<Result<Vec<_>>>().unwrap();
+        meili_snap::snapshot!(meili_snap::json_string!(batches), @"[]");
+
         // tasks
         let tasks = dump.tasks().unwrap().collect::<Result<Vec<_>>>().unwrap();
         let (tasks, update_files): (Vec<_>, Vec<_>) = tasks.into_iter().unzip();
@@ -539,6 +592,10 @@ pub(crate) mod test {
         insta::assert_snapshot!(dump.date().unwrap(), @"2022-10-07 11:39:03.709153554 +00:00:00");
         assert_eq!(dump.instance_uid().unwrap(), None);
 
+        // batches didn't exists at the time
+        let batches = dump.batches().unwrap().collect::<Result<Vec<_>>>().unwrap();
+        meili_snap::snapshot!(meili_snap::json_string!(batches), @"[]");
+
         // tasks
         let tasks = dump.tasks().unwrap().collect::<Result<Vec<_>>>().unwrap();
         let (tasks, update_files): (Vec<_>, Vec<_>) = tasks.into_iter().unzip();
@@ -632,6 +689,10 @@ pub(crate) mod test {
         insta::assert_snapshot!(dump.date().unwrap(), @"2022-10-09 20:27:59.904096267 +00:00:00");
         assert_eq!(dump.instance_uid().unwrap(), None);
 
+        // batches didn't exists at the time
+        let batches = dump.batches().unwrap().collect::<Result<Vec<_>>>().unwrap();
+        meili_snap::snapshot!(meili_snap::json_string!(batches), @"[]");
+
         // tasks
         let tasks = dump.tasks().unwrap().collect::<Result<Vec<_>>>().unwrap();
         let (tasks, update_files): (Vec<_>, Vec<_>) = tasks.into_iter().unzip();
@@ -725,6 +786,10 @@ pub(crate) mod test {
         insta::assert_snapshot!(dump.date().unwrap(), @"2023-01-30 16:26:09.247261 +00:00:00");
         assert_eq!(dump.instance_uid().unwrap(), None);
 
+        // batches didn't exists at the time
+        let batches = dump.batches().unwrap().collect::<Result<Vec<_>>>().unwrap();
+        meili_snap::snapshot!(meili_snap::json_string!(batches), @"[]");
+
         // tasks
         let tasks = dump.tasks().unwrap().collect::<Result<Vec<_>>>().unwrap();
         let (tasks, update_files): (Vec<_>, Vec<_>) = tasks.into_iter().unzip();
@@ -801,6 +866,10 @@ pub(crate) mod test {
         assert_eq!(dump.date(), None);
         assert_eq!(dump.instance_uid().unwrap(), None);
 
+        // batches didn't exists at the time
+        let batches = dump.batches().unwrap().collect::<Result<Vec<_>>>().unwrap();
+        meili_snap::snapshot!(meili_snap::json_string!(batches), @"[]");
+
         // tasks
         let tasks = dump.tasks().unwrap().collect::<Result<Vec<_>>>().unwrap();
         let (tasks, update_files): (Vec<_>, Vec<_>) = tasks.into_iter().unzip();

crates/dump/src/reader/snapshots/dump__reader__test__import_dump_v6_with_vectors-5.snap

@@ -1,5 +1,5 @@
 ---
-source: dump/src/reader/mod.rs
+source: crates/dump/src/reader/mod.rs
 expression: vector_index.settings().unwrap()
 ---
 {
@@ -49,6 +49,7 @@ expression: vector_index.settings().unwrap()
       "source": "huggingFace",
       "model": "BAAI/bge-base-en-v1.5",
       "revision": "617ca489d9e86b49b8167676d8220688b99db36e",
+      "pooling": "forceMean",
       "documentTemplate": "{% for field in fields %} {{ field.name }}: {{ field.value }}\n{% endfor %}"
     }
   },

crates/dump/src/reader/v6/mod.rs

@@ -3,6 +3,7 @@ use std::io::{BufRead, BufReader, ErrorKind};
 use std::path::Path;
 
 pub use meilisearch_types::milli;
+use meilisearch_types::milli::vector::hf::OverridePooling;
 use tempfile::TempDir;
 use time::OffsetDateTime;
 use tracing::debug;
@@ -18,8 +19,10 @@ pub type Checked = meilisearch_types::settings::Checked;
 pub type Unchecked = meilisearch_types::settings::Unchecked;
 
 pub type Task = crate::TaskDump;
+pub type Batch = meilisearch_types::batches::Batch;
 pub type Key = meilisearch_types::keys::Key;
 pub type RuntimeTogglableFeatures = meilisearch_types::features::RuntimeTogglableFeatures;
+pub type Network = meilisearch_types::features::Network;
 
 // ===== Other types to clarify the code of the compat module
 // everything related to the tasks
@@ -43,13 +46,17 @@ pub type ResponseError = meilisearch_types::error::ResponseError;
 pub type Code = meilisearch_types::error::Code;
 pub type RankingRuleView = meilisearch_types::settings::RankingRuleView;
 
+pub type FilterableAttributesRule = meilisearch_types::milli::FilterableAttributesRule;
+
 pub struct V6Reader {
     dump: TempDir,
     instance_uid: Option<Uuid>,
     metadata: Metadata,
     tasks: BufReader<File>,
+    batches: Option<BufReader<File>>,
     keys: BufReader<File>,
     features: Option<RuntimeTogglableFeatures>,
+    network: Option<Network>,
 }
 
 impl V6Reader {
@@ -77,13 +84,38 @@ impl V6Reader {
         } else {
             None
         };
+        let batches = match File::open(dump.path().join("batches").join("queue.jsonl")) {
+            Ok(file) => Some(BufReader::new(file)),
+            // The batch file was only introduced during the v1.13, anything prior to that won't have batches
+            Err(err) if err.kind() == ErrorKind::NotFound => None,
+            Err(e) => return Err(e.into()),
+        };
+
+        let network_file = match fs::read(dump.path().join("network.json")) {
+            Ok(network_file) => Some(network_file),
+            Err(error) => match error.kind() {
+                // Allows the file to be missing, this will only result in all experimental features disabled.
+                ErrorKind::NotFound => {
+                    debug!("`network.json` not found in dump");
+                    None
+                }
+                _ => return Err(error.into()),
+            },
+        };
+        let network = if let Some(network_file) = network_file {
+            Some(serde_json::from_reader(&*network_file)?)
+        } else {
+            None
+        };
 
         Ok(V6Reader {
             metadata: serde_json::from_reader(&*meta_file)?,
             instance_uid,
             tasks: BufReader::new(File::open(dump.path().join("tasks").join("queue.jsonl"))?),
+            batches,
             keys: BufReader::new(File::open(dump.path().join("keys.jsonl"))?),
             features,
+            network,
             dump,
         })
     }
@@ -124,7 +156,7 @@ impl V6Reader {
         &mut self,
     ) -> Box<dyn Iterator<Item = Result<(Task, Option<Box<super::UpdateFile>>)>> + '_> {
         Box::new((&mut self.tasks).lines().map(|line| -> Result<_> {
-            let task: Task = serde_json::from_str(&line?).unwrap();
+            let task: Task = serde_json::from_str(&line?)?;
 
             let update_file_path = self
                 .dump
@@ -136,8 +168,7 @@ impl V6Reader {
             if update_file_path.exists() {
                 Ok((
                     task,
-                    Some(Box::new(UpdateFile::new(&update_file_path).unwrap())
-                        as Box<super::UpdateFile>),
+                    Some(Box::new(UpdateFile::new(&update_file_path)?) as Box<super::UpdateFile>),
                 ))
             } else {
                 Ok((task, None))
@@ -145,6 +176,16 @@ impl V6Reader {
         }))
     }
 
+    pub fn batches(&mut self) -> Box<dyn Iterator<Item = Result<Batch>> + '_> {
+        match self.batches.as_mut() {
+            Some(batches) => Box::new((batches).lines().map(|line| -> Result<_> {
+                let batch = serde_json::from_str(&line?)?;
+                Ok(batch)
+            })),
+            None => Box::new(std::iter::empty()) as Box<dyn Iterator<Item = Result<Batch>> + '_>,
+        }
+    }
+
     pub fn keys(&mut self) -> Box<dyn Iterator<Item = Result<Key>> + '_> {
         Box::new(
             (&mut self.keys).lines().map(|line| -> Result<_> { Ok(serde_json::from_str(&line?)?) }),
@@ -154,6 +195,10 @@ impl V6Reader {
     pub fn features(&self) -> Option<RuntimeTogglableFeatures> {
         self.features
     }
+
+    pub fn network(&self) -> Option<&Network> {
+        self.network.as_ref()
+    }
 }
 
 pub struct UpdateFile {
@@ -210,7 +255,29 @@ impl V6IndexReader {
     }
 
     pub fn settings(&mut self) -> Result<Settings<Checked>> {
-        let settings: Settings<Unchecked> = serde_json::from_reader(&mut self.settings)?;
+        let mut settings: Settings<Unchecked> = serde_json::from_reader(&mut self.settings)?;
+        patch_embedders(&mut settings);
         Ok(settings.check())
     }
 }
+
+fn patch_embedders(settings: &mut Settings<Unchecked>) {
+    if let Setting::Set(embedders) = &mut settings.embedders {
+        for settings in embedders.values_mut() {
+            let Setting::Set(settings) = &mut settings.inner else {
+                continue;
+            };
+            if settings.source != Setting::Set(milli::vector::settings::EmbedderSource::HuggingFace)
+            {
+                continue;
+            }
+            settings.pooling = match settings.pooling {
+                Setting::Set(pooling) => Setting::Set(pooling),
+                // if the pooling for a hugging face embedder is not set, force it to `forceMean`
+                // for backward compatibility with v1.13
+                // dumps created in v1.14 and up will have the setting set for hugging face embedders
+                Setting::Reset | Setting::NotSet => Setting::Set(OverridePooling::ForceMean),
+            };
+        }
+    }
+}

crates/dump/src/writer.rs

@@ -4,7 +4,8 @@ use std::path::PathBuf;
 
 use flate2::write::GzEncoder;
 use flate2::Compression;
-use meilisearch_types::features::RuntimeTogglableFeatures;
+use meilisearch_types::batches::Batch;
+use meilisearch_types::features::{Network, RuntimeTogglableFeatures};
 use meilisearch_types::keys::Key;
 use meilisearch_types::settings::{Checked, Settings};
 use serde_json::{Map, Value};
@@ -54,6 +55,10 @@ impl DumpWriter {
         TaskWriter::new(self.dir.path().join("tasks"))
     }
 
+    pub fn create_batches_queue(&self) -> Result<BatchWriter> {
+        BatchWriter::new(self.dir.path().join("batches"))
+    }
+
     pub fn create_experimental_features(&self, features: RuntimeTogglableFeatures) -> Result<()> {
         Ok(std::fs::write(
             self.dir.path().join("experimental-features.json"),
@@ -61,6 +66,10 @@ impl DumpWriter {
         )?)
     }
 
+    pub fn create_network(&self, network: Network) -> Result<()> {
+        Ok(std::fs::write(self.dir.path().join("network.json"), serde_json::to_string(&network)?)?)
+    }
+
     pub fn persist_to(self, mut writer: impl Write) -> Result<()> {
         let gz_encoder = GzEncoder::new(&mut writer, Compression::default());
         let mut tar_encoder = tar::Builder::new(gz_encoder);
@@ -84,7 +93,7 @@ impl KeyWriter {
     }
 
     pub fn push_key(&mut self, key: &Key) -> Result<()> {
-        self.keys.write_all(&serde_json::to_vec(key)?)?;
+        serde_json::to_writer(&mut self.keys, &key)?;
         self.keys.write_all(b"\n")?;
         Ok(())
     }
@@ -114,7 +123,7 @@ impl TaskWriter {
     /// Pushes tasks in the dump.
     /// If the tasks has an associated `update_file` it'll use the `task_id` as its name.
     pub fn push_task(&mut self, task: &TaskDump) -> Result<UpdateFile> {
-        self.queue.write_all(&serde_json::to_vec(task)?)?;
+        serde_json::to_writer(&mut self.queue, &task)?;
         self.queue.write_all(b"\n")?;
 
         Ok(UpdateFile::new(self.update_files.join(format!("{}.jsonl", task.uid))))
@@ -126,6 +135,30 @@ impl TaskWriter {
     }
 }
 
+pub struct BatchWriter {
+    queue: BufWriter<File>,
+}
+
+impl BatchWriter {
+    pub(crate) fn new(path: PathBuf) -> Result<Self> {
+        std::fs::create_dir(&path)?;
+        let queue = File::create(path.join("queue.jsonl"))?;
+        Ok(BatchWriter { queue: BufWriter::new(queue) })
+    }
+
+    /// Pushes batches in the dump.
+    pub fn push_batch(&mut self, batch: &Batch) -> Result<()> {
+        serde_json::to_writer(&mut self.queue, &batch)?;
+        self.queue.write_all(b"\n")?;
+        Ok(())
+    }
+
+    pub fn flush(mut self) -> Result<()> {
+        self.queue.flush()?;
+        Ok(())
+    }
+}
+
 pub struct UpdateFile {
     path: PathBuf,
     writer: Option<BufWriter<File>>,
@@ -137,8 +170,8 @@ impl UpdateFile {
     }
 
     pub fn push_document(&mut self, document: &Document) -> Result<()> {
-        if let Some(writer) = self.writer.as_mut() {
-            writer.write_all(&serde_json::to_vec(document)?)?;
+        if let Some(mut writer) = self.writer.as_mut() {
+            serde_json::to_writer(&mut writer, &document)?;
             writer.write_all(b"\n")?;
         } else {
             let file = File::create(&self.path).unwrap();
@@ -205,8 +238,8 @@ pub(crate) mod test {
     use super::*;
     use crate::reader::Document;
     use crate::test::{
-        create_test_api_keys, create_test_documents, create_test_dump, create_test_instance_uid,
-        create_test_settings, create_test_tasks,
+        create_test_api_keys, create_test_batches, create_test_documents, create_test_dump,
+        create_test_instance_uid, create_test_settings, create_test_tasks,
     };
 
     fn create_directory_hierarchy(dir: &Path) -> String {
@@ -281,8 +314,10 @@ pub(crate) mod test {
         let dump_path = dump.path();
 
         // ==== checking global file hierarchy (we want to be sure there isn't too many files or too few)
-        insta::assert_snapshot!(create_directory_hierarchy(dump_path), @r###"
+        insta::assert_snapshot!(create_directory_hierarchy(dump_path), @r"
         .
+        ├---- batches/
+        │    └---- queue.jsonl
         ├---- indexes/
         │    └---- doggos/
         │    │    ├---- documents.jsonl
@@ -295,8 +330,9 @@ pub(crate) mod test {
         ├---- experimental-features.json
         ├---- instance_uid.uuid
         ├---- keys.jsonl
-        └---- metadata.json
-        "###);
+        ├---- metadata.json
+        └---- network.json
+        ");
 
         // ==== checking the top level infos
         let metadata = fs::read_to_string(dump_path.join("metadata.json")).unwrap();
@@ -349,6 +385,16 @@ pub(crate) mod test {
             }
         }
 
+        // ==== checking the batch queue
+        let batches_queue = fs::read_to_string(dump_path.join("batches/queue.jsonl")).unwrap();
+        for (batch, expected) in batches_queue.lines().zip(create_test_batches()) {
+            let mut batch = serde_json::from_str::<Batch>(batch).unwrap();
+            if batch.details.settings == Some(Box::new(Settings::<Unchecked>::default())) {
+                batch.details.settings = None;
+            }
+            assert_eq!(batch, expected, "{batch:#?}{expected:#?}");
+        }
+
         // ==== checking the keys
         let keys = fs::read_to_string(dump_path.join("keys.jsonl")).unwrap();
         for (key, expected) in keys.lines().zip(create_test_api_keys()) {

crates/filter-parser/src/condition.rs

@@ -30,6 +30,25 @@ pub enum Condition<'a> {
     StartsWith { keyword: Token<'a>, word: Token<'a> },
 }
 
+impl Condition<'_> {
+    pub fn operator(&self) -> &str {
+        match self {
+            Condition::GreaterThan(_) => ">",
+            Condition::GreaterThanOrEqual(_) => ">=",
+            Condition::Equal(_) => "=",
+            Condition::NotEqual(_) => "!=",
+            Condition::Null => "IS NULL",
+            Condition::Empty => "IS EMPTY",
+            Condition::Exists => "EXISTS",
+            Condition::LowerThan(_) => "<",
+            Condition::LowerThanOrEqual(_) => "<=",
+            Condition::Between { .. } => "TO",
+            Condition::Contains { .. } => "CONTAINS",
+            Condition::StartsWith { .. } => "STARTS WITH",
+        }
+    }
+}
+
 /// condition      = value ("==" | ">" ...) value
 pub fn parse_condition(input: Span) -> IResult<FilterCondition> {
     let operator = alt((tag("<="), tag(">="), tag("!="), tag("<"), tag(">"), tag("=")));

crates/index-scheduler/Cargo.toml

@@ -29,7 +29,7 @@ page_size = "0.6.0"
 rayon = "1.10.0"
 roaring = { version = "0.10.10", features = ["serde"] }
 serde = { version = "1.0.217", features = ["derive"] }
-serde_json = { version = "1.0.135", features = ["preserve_order"] }
+serde_json = { version = "1.0.138", features = ["preserve_order"] }
 synchronoise = "1.0.1"
 tempfile = "3.15.0"
 thiserror = "2.0.9"

crates/index-scheduler/src/dump.rs

@@ -2,6 +2,7 @@ use std::collections::HashMap;
 use std::io;
 
 use dump::{KindDump, TaskDump, UpdateFile};
+use meilisearch_types::batches::{Batch, BatchId};
 use meilisearch_types::heed::RwTxn;
 use meilisearch_types::milli;
 use meilisearch_types::tasks::{Kind, KindWithContent, Status, Task};
@@ -14,9 +15,15 @@ pub struct Dump<'a> {
     index_scheduler: &'a IndexScheduler,
     wtxn: RwTxn<'a>,
 
+    batch_to_task_mapping: HashMap<BatchId, RoaringBitmap>,
+
     indexes: HashMap<String, RoaringBitmap>,
     statuses: HashMap<Status, RoaringBitmap>,
     kinds: HashMap<Kind, RoaringBitmap>,
+
+    batch_indexes: HashMap<String, RoaringBitmap>,
+    batch_statuses: HashMap<Status, RoaringBitmap>,
+    batch_kinds: HashMap<Kind, RoaringBitmap>,
 }
 
 impl<'a> Dump<'a> {
@@ -27,12 +34,72 @@ impl<'a> Dump<'a> {
         Ok(Dump {
             index_scheduler,
             wtxn,
+            batch_to_task_mapping: HashMap::new(),
             indexes: HashMap::new(),
             statuses: HashMap::new(),
             kinds: HashMap::new(),
+            batch_indexes: HashMap::new(),
+            batch_statuses: HashMap::new(),
+            batch_kinds: HashMap::new(),
         })
     }
 
+    /// Register a new batch coming from a dump in the scheduler.
+    /// By taking a mutable ref we're pretty sure no one will ever import a dump while actix is running.
+    pub fn register_dumped_batch(&mut self, batch: Batch) -> Result<()> {
+        self.index_scheduler.queue.batches.all_batches.put(&mut self.wtxn, &batch.uid, &batch)?;
+        if let Some(enqueued_at) = batch.enqueued_at {
+            utils::insert_task_datetime(
+                &mut self.wtxn,
+                self.index_scheduler.queue.batches.enqueued_at,
+                enqueued_at.earliest,
+                batch.uid,
+            )?;
+            utils::insert_task_datetime(
+                &mut self.wtxn,
+                self.index_scheduler.queue.batches.enqueued_at,
+                enqueued_at.oldest,
+                batch.uid,
+            )?;
+        }
+        utils::insert_task_datetime(
+            &mut self.wtxn,
+            self.index_scheduler.queue.batches.started_at,
+            batch.started_at,
+            batch.uid,
+        )?;
+        if let Some(finished_at) = batch.finished_at {
+            utils::insert_task_datetime(
+                &mut self.wtxn,
+                self.index_scheduler.queue.batches.finished_at,
+                finished_at,
+                batch.uid,
+            )?;
+        }
+
+        for index in batch.stats.index_uids.keys() {
+            match self.batch_indexes.get_mut(index) {
+                Some(bitmap) => {
+                    bitmap.insert(batch.uid);
+                }
+                None => {
+                    let mut bitmap = RoaringBitmap::new();
+                    bitmap.insert(batch.uid);
+                    self.batch_indexes.insert(index.to_string(), bitmap);
+                }
+            };
+        }
+
+        for status in batch.stats.status.keys() {
+            self.batch_statuses.entry(*status).or_default().insert(batch.uid);
+        }
+        for kind in batch.stats.types.keys() {
+            self.batch_kinds.entry(*kind).or_default().insert(batch.uid);
+        }
+
+        Ok(())
+    }
+
     /// Register a new task coming from a dump in the scheduler.
     /// By taking a mutable ref we're pretty sure no one will ever import a dump while actix is running.
     pub fn register_dumped_task(
@@ -149,6 +216,9 @@ impl<'a> Dump<'a> {
         };
 
         self.index_scheduler.queue.tasks.all_tasks.put(&mut self.wtxn, &task.uid, &task)?;
+        if let Some(batch_id) = task.batch_uid {
+            self.batch_to_task_mapping.entry(batch_id).or_default().insert(task.uid);
+        }
 
         for index in task.indexes() {
             match self.indexes.get_mut(index) {
@@ -198,6 +268,14 @@ impl<'a> Dump<'a> {
 
     /// Commit all the changes and exit the importing dump state
     pub fn finish(mut self) -> Result<()> {
+        for (batch_id, task_ids) in self.batch_to_task_mapping {
+            self.index_scheduler.queue.batch_to_tasks_mapping.put(
+                &mut self.wtxn,
+                &batch_id,
+                &task_ids,
+            )?;
+        }
+
         for (index, bitmap) in self.indexes {
             self.index_scheduler.queue.tasks.index_tasks.put(&mut self.wtxn, &index, &bitmap)?;
         }
@@ -208,6 +286,16 @@ impl<'a> Dump<'a> {
             self.index_scheduler.queue.tasks.put_kind(&mut self.wtxn, kind, &bitmap)?;
         }
 
+        for (index, bitmap) in self.batch_indexes {
+            self.index_scheduler.queue.batches.index_tasks.put(&mut self.wtxn, &index, &bitmap)?;
+        }
+        for (status, bitmap) in self.batch_statuses {
+            self.index_scheduler.queue.batches.put_status(&mut self.wtxn, status, &bitmap)?;
+        }
+        for (kind, bitmap) in self.batch_kinds {
+            self.index_scheduler.queue.batches.put_kind(&mut self.wtxn, kind, &bitmap)?;
+        }
+
         self.wtxn.commit()?;
         self.index_scheduler.scheduler.wake_up.signal();
 

crates/index-scheduler/src/error.rs

@@ -109,6 +109,8 @@ pub enum Error {
     InvalidIndexUid { index_uid: String },
     #[error("Task `{0}` not found.")]
     TaskNotFound(TaskId),
+    #[error("Task `{0}` does not contain any documents. Only `documentAdditionOrUpdate` tasks with the statuses `enqueued` or `processing` contain documents")]
+    TaskFileNotFound(TaskId),
     #[error("Batch `{0}` not found.")]
     BatchNotFound(BatchId),
     #[error("Query parameters to filter the tasks to delete are missing. Available query parameters are: `uids`, `indexUids`, `statuses`, `types`, `canceledBy`, `beforeEnqueuedAt`, `afterEnqueuedAt`, `beforeStartedAt`, `afterStartedAt`, `beforeFinishedAt`, `afterFinishedAt`.")]
@@ -127,8 +129,8 @@ pub enum Error {
         _ => format!("{error}")
     })]
     Milli { error: milli::Error, index_uid: Option<String> },
-    #[error("An unexpected crash occurred when processing the task.")]
-    ProcessBatchPanicked,
+    #[error("An unexpected crash occurred when processing the task: {0}")]
+    ProcessBatchPanicked(String),
     #[error(transparent)]
     FileStore(#[from] file_store::Error),
     #[error(transparent)]
@@ -189,6 +191,7 @@ impl Error {
             | Error::InvalidTaskCanceledBy { .. }
             | Error::InvalidIndexUid { .. }
             | Error::TaskNotFound(_)
+            | Error::TaskFileNotFound(_)
             | Error::BatchNotFound(_)
             | Error::TaskDeletionWithEmptyQuery
             | Error::TaskCancelationWithEmptyQuery
@@ -196,7 +199,7 @@ impl Error {
             | Error::Dump(_)
             | Error::Heed(_)
             | Error::Milli { .. }
-            | Error::ProcessBatchPanicked
+            | Error::ProcessBatchPanicked(_)
             | Error::FileStore(_)
             | Error::IoError(_)
             | Error::Persist(_)
@@ -250,6 +253,7 @@ impl ErrorCode for Error {
             Error::InvalidTaskCanceledBy { .. } => Code::InvalidTaskCanceledBy,
             Error::InvalidIndexUid { .. } => Code::InvalidIndexUid,
             Error::TaskNotFound(_) => Code::TaskNotFound,
+            Error::TaskFileNotFound(_) => Code::TaskFileNotFound,
             Error::BatchNotFound(_) => Code::BatchNotFound,
             Error::TaskDeletionWithEmptyQuery => Code::MissingTaskFilters,
             Error::TaskCancelationWithEmptyQuery => Code::MissingTaskFilters,
@@ -257,7 +261,7 @@ impl ErrorCode for Error {
             Error::NoSpaceLeftInTaskQueue => Code::NoSpaceLeftOnDevice,
             Error::Dump(e) => e.error_code(),
             Error::Milli { error, .. } => error.error_code(),
-            Error::ProcessBatchPanicked => Code::Internal,
+            Error::ProcessBatchPanicked(_) => Code::Internal,
             Error::Heed(e) => e.error_code(),
             Error::HeedTransaction(e) => e.error_code(),
             Error::FileStore(e) => e.error_code(),

crates/index-scheduler/src/features.rs

@@ -1,6 +1,6 @@
 use std::sync::{Arc, RwLock};
 
-use meilisearch_types::features::{InstanceTogglableFeatures, RuntimeTogglableFeatures};
+use meilisearch_types::features::{InstanceTogglableFeatures, Network, RuntimeTogglableFeatures};
 use meilisearch_types::heed::types::{SerdeJson, Str};
 use meilisearch_types::heed::{Database, Env, RwTxn};
 
@@ -14,10 +14,16 @@ mod db_name {
     pub const EXPERIMENTAL_FEATURES: &str = "experimental-features";
 }
 
+mod db_keys {
+    pub const EXPERIMENTAL_FEATURES: &str = "experimental-features";
+    pub const NETWORK: &str = "network";
+}
+
 #[derive(Clone)]
 pub(crate) struct FeatureData {
     persisted: Database<Str, SerdeJson<RuntimeTogglableFeatures>>,
     runtime: Arc<RwLock<RuntimeTogglableFeatures>>,
+    network: Arc<RwLock<Network>>,
 }
 
 #[derive(Debug, Clone, Copy)]
@@ -86,6 +92,32 @@ impl RoFeatures {
             .into())
         }
     }
+
+    pub fn check_network(&self, disabled_action: &'static str) -> Result<()> {
+        if self.runtime.network {
+            Ok(())
+        } else {
+            Err(FeatureNotEnabledError {
+                disabled_action,
+                feature: "network",
+                issue_link: "https://github.com/orgs/meilisearch/discussions/805",
+            }
+            .into())
+        }
+    }
+
+    pub fn check_get_task_documents_route(&self) -> Result<()> {
+        if self.runtime.get_task_documents_route {
+            Ok(())
+        } else {
+            Err(FeatureNotEnabledError {
+                disabled_action: "Getting the documents of an enqueued task",
+                feature: "get task documents route",
+                issue_link: "https://github.com/orgs/meilisearch/discussions/808",
+            }
+            .into())
+        }
+    }
 }
 
 impl FeatureData {
@@ -102,7 +134,7 @@ impl FeatureData {
             env.create_database(wtxn, Some(db_name::EXPERIMENTAL_FEATURES))?;
 
         let persisted_features: RuntimeTogglableFeatures =
-            runtime_features_db.get(wtxn, db_name::EXPERIMENTAL_FEATURES)?.unwrap_or_default();
+            runtime_features_db.get(wtxn, db_keys::EXPERIMENTAL_FEATURES)?.unwrap_or_default();
         let InstanceTogglableFeatures { metrics, logs_route, contains_filter } = instance_features;
         let runtime = Arc::new(RwLock::new(RuntimeTogglableFeatures {
             metrics: metrics || persisted_features.metrics,
@@ -111,7 +143,14 @@ impl FeatureData {
             ..persisted_features
         }));
 
-        Ok(Self { persisted: runtime_features_db, runtime })
+        let network_db = runtime_features_db.remap_data_type::<SerdeJson<Network>>();
+        let network: Network = network_db.get(wtxn, db_keys::NETWORK)?.unwrap_or_default();
+
+        Ok(Self {
+            persisted: runtime_features_db,
+            runtime,
+            network: Arc::new(RwLock::new(network)),
+        })
     }
 
     pub fn put_runtime_features(
@@ -119,7 +158,7 @@ impl FeatureData {
         mut wtxn: RwTxn,
         features: RuntimeTogglableFeatures,
     ) -> Result<()> {
-        self.persisted.put(&mut wtxn, db_name::EXPERIMENTAL_FEATURES, &features)?;
+        self.persisted.put(&mut wtxn, db_keys::EXPERIMENTAL_FEATURES, &features)?;
         wtxn.commit()?;
 
         // safe to unwrap, the lock will only fail if:
@@ -140,4 +179,21 @@ impl FeatureData {
     pub fn features(&self) -> RoFeatures {
         RoFeatures::new(self)
     }
+
+    pub fn put_network(&self, mut wtxn: RwTxn, new_network: Network) -> Result<()> {
+        self.persisted.remap_data_type::<SerdeJson<Network>>().put(
+            &mut wtxn,
+            db_keys::NETWORK,
+            &new_network,
+        )?;
+        wtxn.commit()?;
+
+        let mut network = self.network.write().unwrap();
+        *network = new_network;
+        Ok(())
+    }
+
+    pub fn network(&self) -> Network {
+        Network::clone(&*self.network.read().unwrap())
+    }
 }

crates/index-scheduler/src/index_mapper/index_map.rs

@@ -1,5 +1,7 @@
 use std::collections::BTreeMap;
+use std::env::VarError;
 use std::path::Path;
+use std::str::FromStr;
 use std::time::Duration;
 
 use meilisearch_types::heed::{EnvClosingEvent, EnvFlags, EnvOpenOptions};
@@ -304,7 +306,18 @@ fn create_or_open_index(
 ) -> Result<Index> {
     let mut options = EnvOpenOptions::new();
     options.map_size(clamp_to_page_size(map_size));
-    options.max_readers(1024);
+
+    // You can find more details about this experimental
+    // environment variable on the following GitHub discussion:
+    // <https://github.com/orgs/meilisearch/discussions/806>
+    let max_readers = match std::env::var("MEILI_EXPERIMENTAL_INDEX_MAX_READERS") {
+        Ok(value) => u32::from_str(&value).unwrap(),
+        Err(VarError::NotPresent) => 1024,
+        Err(VarError::NotUnicode(value)) => panic!(
+            "Invalid unicode for the `MEILI_EXPERIMENTAL_INDEX_MAX_READERS` env var: {value:?}"
+        ),
+    };
+    options.max_readers(max_readers);
     if enable_mdb_writemap {
         unsafe { options.flags(EnvFlags::WRITE_MAP) };
     }

crates/index-scheduler/src/index_mapper/mod.rs

@@ -6,6 +6,7 @@ use std::{fs, thread};
 use meilisearch_types::heed::types::{SerdeJson, Str};
 use meilisearch_types::heed::{Database, Env, RoTxn, RwTxn};
 use meilisearch_types::milli;
+use meilisearch_types::milli::database_stats::DatabaseStats;
 use meilisearch_types::milli::update::IndexerConfig;
 use meilisearch_types::milli::{FieldDistribution, Index};
 use serde::{Deserialize, Serialize};
@@ -98,14 +99,25 @@ pub enum IndexStatus {
 /// The statistics that can be computed from an `Index` object.
 #[derive(Serialize, Deserialize, Debug)]
 pub struct IndexStats {
-    /// Number of documents in the index.
-    pub number_of_documents: u64,
+    /// Stats of the documents database.
+    #[serde(default)]
+    pub documents_database_stats: DatabaseStats,
+
+    #[serde(default, skip_serializing)]
+    pub number_of_documents: Option<u64>,
+
     /// Size taken up by the index' DB, in bytes.
     ///
     /// This includes the size taken by both the used and free pages of the DB, and as the free pages
     /// are not returned to the disk after a deletion, this number is typically larger than
     /// `used_database_size` that only includes the size of the used pages.
     pub database_size: u64,
+    /// Number of embeddings in the index.
+    /// Option: retrocompatible with the stats of the pre-v1.13.0 versions of meilisearch
+    pub number_of_embeddings: Option<u64>,
+    /// Number of embedded documents in the index.
+    /// Option: retrocompatible with the stats of the pre-v1.13.0 versions of meilisearch
+    pub number_of_embedded_documents: Option<u64>,
     /// Size taken by the used pages of the index' DB, in bytes.
     ///
     /// As the DB backend does not return to the disk the pages that are not currently used by the DB,
@@ -130,8 +142,12 @@ impl IndexStats {
     ///
     /// - rtxn: a RO transaction for the index, obtained from `Index::read_txn()`.
     pub fn new(index: &Index, rtxn: &RoTxn) -> milli::Result<Self> {
+        let arroy_stats = index.arroy_stats(rtxn)?;
         Ok(IndexStats {
-            number_of_documents: index.number_of_documents(rtxn)?,
+            number_of_embeddings: Some(arroy_stats.number_of_embeddings),
+            number_of_embedded_documents: Some(arroy_stats.documents.len()),
+            documents_database_stats: index.documents_stats(rtxn)?.unwrap_or_default(),
+            number_of_documents: None,
             database_size: index.on_disk_size()?,
             used_database_size: index.used_size()?,
             primary_key: index.primary_key(rtxn)?.map(|s| s.to_string()),

crates/index-scheduler/src/insta_snapshot.rs

@@ -1,7 +1,7 @@
 use std::collections::BTreeSet;
 use std::fmt::Write;
 
-use meilisearch_types::batches::Batch;
+use meilisearch_types::batches::{Batch, BatchEnqueuedAt, BatchStats};
 use meilisearch_types::heed::types::{SerdeBincode, SerdeJson, Str};
 use meilisearch_types::heed::{Database, RoTxn};
 use meilisearch_types::milli::{CboRoaringBitmapCodec, RoaringBitmapCodec, BEU32};
@@ -341,14 +341,23 @@ pub fn snapshot_canceled_by(rtxn: &RoTxn, db: Database<BEU32, RoaringBitmapCodec
 
 pub fn snapshot_batch(batch: &Batch) -> String {
     let mut snap = String::new();
-    let Batch { uid, details, stats, started_at, finished_at, progress: _ } = batch;
+    let Batch { uid, details, stats, started_at, finished_at, progress: _, enqueued_at } = batch;
+    let stats = BatchStats {
+        progress_trace: Default::default(),
+        write_channel_congestion: None,
+        ..stats.clone()
+    };
     if let Some(finished_at) = finished_at {
         assert!(finished_at > started_at);
     }
+    let BatchEnqueuedAt { earliest, oldest } = enqueued_at.unwrap();
+    assert!(*started_at > earliest);
+    assert!(earliest >= oldest);
+
     snap.push('{');
     snap.push_str(&format!("uid: {uid}, "));
     snap.push_str(&format!("details: {}, ", serde_json::to_string(details).unwrap()));
-    snap.push_str(&format!("stats: {}, ", serde_json::to_string(stats).unwrap()));
+    snap.push_str(&format!("stats: {}, ", serde_json::to_string(&stats).unwrap()));
     snap.push('}');
     snap
 }
@@ -361,7 +370,8 @@ pub fn snapshot_index_mapper(rtxn: &RoTxn, mapper: &IndexMapper) -> String {
         let stats = mapper.stats_of(rtxn, &name).unwrap();
         s.push_str(&format!(
             "{name}: {{ number_of_documents: {}, field_distribution: {:?} }}\n",
-            stats.number_of_documents, stats.field_distribution
+            stats.documents_database_stats.number_of_entries(),
+            stats.field_distribution
         ));
     }
 

crates/index-scheduler/src/lib.rs

@@ -33,7 +33,7 @@ mod test_utils;
 pub mod upgrade;
 mod utils;
 pub mod uuid_codec;
-mod versioning;
+pub mod versioning;
 
 pub type Result<T, E = Error> = std::result::Result<T, E>;
 pub type TaskId = u32;
@@ -51,7 +51,7 @@ pub use features::RoFeatures;
 use flate2::bufread::GzEncoder;
 use flate2::Compression;
 use meilisearch_types::batches::Batch;
-use meilisearch_types::features::{InstanceTogglableFeatures, RuntimeTogglableFeatures};
+use meilisearch_types::features::{InstanceTogglableFeatures, Network, RuntimeTogglableFeatures};
 use meilisearch_types::heed::byteorder::BE;
 use meilisearch_types::heed::types::I128;
 use meilisearch_types::heed::{self, Env, RoTxn};
@@ -770,7 +770,16 @@ impl IndexScheduler {
         Ok(())
     }
 
-    // TODO: consider using a type alias or a struct embedder/template
+    pub fn put_network(&self, network: Network) -> Result<()> {
+        let wtxn = self.env.write_txn().map_err(Error::HeedTransaction)?;
+        self.features.put_network(wtxn, network)?;
+        Ok(())
+    }
+
+    pub fn network(&self) -> Network {
+        self.features.network()
+    }
+
     pub fn embedders(
         &self,
         index_uid: String,

crates/index-scheduler/src/processing.rs

@@ -96,6 +96,7 @@ make_enum_progress! {
         StartTheDumpCreation,
         DumpTheApiKeys,
         DumpTheTasks,
+        DumpTheBatches,
         DumpTheIndexes,
         DumpTheExperimentalFeatures,
         CompressTheDump,

crates/index-scheduler/src/queue/batches.rs

@@ -12,8 +12,8 @@ use time::OffsetDateTime;
 use super::{Query, Queue};
 use crate::processing::ProcessingTasks;
 use crate::utils::{
-    insert_task_datetime, keep_ids_within_datetimes, map_bound, remove_task_datetime,
-    ProcessingBatch,
+    insert_task_datetime, keep_ids_within_datetimes, map_bound,
+    remove_n_tasks_datetime_earlier_than, remove_task_datetime, ProcessingBatch,
 };
 use crate::{Error, Result, BEI128};
 
@@ -181,6 +181,7 @@ impl BatchQueue {
                 stats: batch.stats,
                 started_at: batch.started_at,
                 finished_at: batch.finished_at,
+                enqueued_at: batch.enqueued_at,
             },
         )?;
 
@@ -234,34 +235,25 @@ impl BatchQueue {
         // What we know, though, is that the task date is from before the enqueued_at, and max two timestamps have been written
         // to the DB per batches.
         if let Some(ref old_batch) = old_batch {
-            let started_at = old_batch.started_at.unix_timestamp_nanos();
-
-            // We have either one or two enqueued at to remove
-            let mut exit = old_batch.stats.total_nb_tasks.clamp(0, 2);
-            let mut iterator = self.enqueued_at.rev_iter_mut(wtxn)?;
-            while let Some(entry) = iterator.next() {
-                let (key, mut value) = entry?;
-                if key > started_at {
-                    continue;
-                }
-                if value.remove(old_batch.uid) {
-                    exit = exit.saturating_sub(1);
-                    // Safe because the key and value are owned
-                    unsafe {
-                        iterator.put_current(&key, &value)?;
-                    }
-                    if exit == 0 {
-                        break;
-                    }
-                }
+            if let Some(enqueued_at) = old_batch.enqueued_at {
+                remove_task_datetime(wtxn, self.enqueued_at, enqueued_at.earliest, old_batch.uid)?;
+                remove_task_datetime(wtxn, self.enqueued_at, enqueued_at.oldest, old_batch.uid)?;
+            } else {
+                // If we don't have the enqueued at in the batch it means the database comes from the v1.12
+                // and we still need to find the date by scrolling the database
+                remove_n_tasks_datetime_earlier_than(
+                    wtxn,
+                    self.enqueued_at,
+                    old_batch.started_at,
+                    old_batch.stats.total_nb_tasks.clamp(1, 2) as usize,
+                    old_batch.uid,
+                )?;
             }
         }
-        if let Some(enqueued_at) = batch.oldest_enqueued_at {
-            insert_task_datetime(wtxn, self.enqueued_at, enqueued_at, batch.uid)?;
-        }
-        if let Some(enqueued_at) = batch.earliest_enqueued_at {
-            insert_task_datetime(wtxn, self.enqueued_at, enqueued_at, batch.uid)?;
-        }
+        // A finished batch MUST contains at least one task and have an enqueued_at
+        let enqueued_at = batch.enqueued_at.as_ref().unwrap();
+        insert_task_datetime(wtxn, self.enqueued_at, enqueued_at.earliest, batch.uid)?;
+        insert_task_datetime(wtxn, self.enqueued_at, enqueued_at.oldest, batch.uid)?;
 
         // Update the started at and finished at
         if let Some(ref old_batch) = old_batch {

crates/index-scheduler/src/queue/batches_test.rs

@@ -102,30 +102,33 @@ fn query_batches_simple() {
         .unwrap();
     assert_eq!(batches.len(), 1);
     batches[0].started_at = OffsetDateTime::UNIX_EPOCH;
+    assert!(batches[0].enqueued_at.is_some());
+    batches[0].enqueued_at = None;
     // Insta cannot snapshot our batches because the batch stats contains an enum as key: https://github.com/mitsuhiko/insta/issues/689
     let batch = serde_json::to_string_pretty(&batches[0]).unwrap();
     snapshot!(batch, @r#"
-        {
-          "uid": 0,
-          "details": {
-            "primaryKey": "mouse"
-          },
-          "stats": {
-            "totalNbTasks": 1,
-            "status": {
-              "processing": 1
-            },
-            "types": {
-              "indexCreation": 1
-            },
-            "indexUids": {
-              "catto": 1
-            }
-          },
-          "startedAt": "1970-01-01T00:00:00Z",
-          "finishedAt": null
+    {
+      "uid": 0,
+      "details": {
+        "primaryKey": "mouse"
+      },
+      "stats": {
+        "totalNbTasks": 1,
+        "status": {
+          "processing": 1
+        },
+        "types": {
+          "indexCreation": 1
+        },
+        "indexUids": {
+          "catto": 1
         }
-        "#);
+      },
+      "startedAt": "1970-01-01T00:00:00Z",
+      "finishedAt": null,
+      "enqueuedAt": null
+    }
+    "#);
 
     let query = Query { statuses: Some(vec![Status::Enqueued]), ..Default::default() };
     let (batches, _) = index_scheduler

crates/index-scheduler/src/queue/mod.rs

@@ -8,6 +8,7 @@ mod tasks_test;
 mod test;
 
 use std::collections::BTreeMap;
+use std::fs::File as StdFile;
 use std::time::Duration;
 
 use file_store::FileStore;
@@ -216,6 +217,11 @@ impl Queue {
         }
     }
 
+    /// Open and returns the task's content File.
+    pub fn update_file(&self, uuid: Uuid) -> file_store::Result<StdFile> {
+        self.file_store.get_update(uuid)
+    }
+
     /// Delete a file from the index scheduler.
     ///
     /// Counterpart to the [`create_update_file`](IndexScheduler::create_update_file) method.

crates/index-scheduler/src/queue/test.rs

@@ -326,7 +326,7 @@ fn test_auto_deletion_of_tasks() {
 fn test_task_queue_is_full() {
     let (index_scheduler, mut handle) = IndexScheduler::test_with_custom_config(vec![], |config| {
         // that's the minimum map size possible
-        config.task_db_size = 1048576;
+        config.task_db_size = 1048576 * 3;
         None
     });
 

crates/index-scheduler/src/scheduler/mod.rs

@@ -166,13 +166,41 @@ impl IndexScheduler {
             let processing_batch = &mut processing_batch;
             let progress = progress.clone();
             std::thread::scope(|s| {
+                let p = progress.clone();
                 let handle = std::thread::Builder::new()
                     .name(String::from("batch-operation"))
                     .spawn_scoped(s, move || {
-                        cloned_index_scheduler.process_batch(batch, processing_batch, progress)
+                        cloned_index_scheduler.process_batch(batch, processing_batch, p)
                     })
                     .unwrap();
-                handle.join().unwrap_or(Err(Error::ProcessBatchPanicked))
+
+                match handle.join() {
+                    Ok(ret) => {
+                        if ret.is_err() {
+                            if let Ok(progress_view) =
+                                serde_json::to_string(&progress.as_progress_view())
+                            {
+                                tracing::warn!("Batch failed while doing: {progress_view}")
+                            }
+                        }
+                        ret
+                    }
+                    Err(panic) => {
+                        if let Ok(progress_view) =
+                            serde_json::to_string(&progress.as_progress_view())
+                        {
+                            tracing::warn!("Batch failed while doing: {progress_view}")
+                        }
+                        let msg = match panic.downcast_ref::<&'static str>() {
+                            Some(s) => *s,
+                            None => match panic.downcast_ref::<String>() {
+                                Some(s) => &s[..],
+                                None => "Box<dyn Any>",
+                            },
+                        };
+                        Err(Error::ProcessBatchPanicked(msg.to_string()))
+                    }
+                }
             })
         };
 
@@ -187,14 +215,16 @@ impl IndexScheduler {
         let mut stop_scheduler_forever = false;
         let mut wtxn = self.env.write_txn().map_err(Error::HeedTransaction)?;
         let mut canceled = RoaringBitmap::new();
+        let mut congestion = None;
 
         match res {
-            Ok(tasks) => {
+            Ok((tasks, cong)) => {
                 #[cfg(test)]
                 self.breakpoint(crate::test_utils::Breakpoint::ProcessBatchSucceeded);
 
                 let (task_progress, task_progress_obj) = AtomicTaskStep::new(tasks.len() as u32);
                 progress.update_progress(task_progress_obj);
+                congestion = cong;
                 let mut success = 0;
                 let mut failure = 0;
                 let mut canceled_by = None;
@@ -311,6 +341,28 @@ impl IndexScheduler {
 
         // We must re-add the canceled task so they're part of the same batch.
         ids |= canceled;
+
+        processing_batch.stats.progress_trace =
+            progress.accumulated_durations().into_iter().map(|(k, v)| (k, v.into())).collect();
+        processing_batch.stats.write_channel_congestion = congestion.map(|congestion| {
+            let mut congestion_info = serde_json::Map::new();
+            congestion_info.insert("attempts".into(), congestion.attempts.into());
+            congestion_info.insert("blocking_attempts".into(), congestion.blocking_attempts.into());
+            congestion_info.insert("blocking_ratio".into(), congestion.congestion_ratio().into());
+            congestion_info
+        });
+
+        if let Some(congestion) = congestion {
+            tracing::debug!(
+                "Channel congestion metrics - Attempts: {}, Blocked attempts: {}  ({:.1}% congestion)",
+                congestion.attempts,
+                congestion.blocking_attempts,
+                congestion.congestion_ratio(),
+            );
+        }
+
+        tracing::debug!("call trace: {:?}", progress.accumulated_durations());
+
         self.queue.write_batch(&mut wtxn, processing_batch, &ids)?;
 
         #[cfg(test)]

crates/index-scheduler/src/scheduler/process_batch.rs

@@ -2,10 +2,10 @@ use std::collections::{BTreeSet, HashMap, HashSet};
 use std::panic::{catch_unwind, AssertUnwindSafe};
 use std::sync::atomic::Ordering;
 
-use meilisearch_types::batches::BatchId;
+use meilisearch_types::batches::{BatchEnqueuedAt, BatchId};
 use meilisearch_types::heed::{RoTxn, RwTxn};
 use meilisearch_types::milli::progress::{Progress, VariableNameStep};
-use meilisearch_types::milli::{self};
+use meilisearch_types::milli::{self, ChannelCongestion};
 use meilisearch_types::tasks::{Details, IndexSwap, KindWithContent, Status, Task};
 use milli::update::Settings as MilliSettings;
 use roaring::RoaringBitmap;
@@ -16,7 +16,10 @@ use crate::processing::{
     InnerSwappingTwoIndexes, SwappingTheIndexes, TaskCancelationProgress, TaskDeletionProgress,
     UpdateIndexProgress,
 };
-use crate::utils::{self, swap_index_uid_in_task, ProcessingBatch};
+use crate::utils::{
+    self, remove_n_tasks_datetime_earlier_than, remove_task_datetime, swap_index_uid_in_task,
+    ProcessingBatch,
+};
 use crate::{Error, IndexScheduler, Result, TaskId};
 
 impl IndexScheduler {
@@ -32,7 +35,7 @@ impl IndexScheduler {
         batch: Batch,
         current_batch: &mut ProcessingBatch,
         progress: Progress,
-    ) -> Result<Vec<Task>> {
+    ) -> Result<(Vec<Task>, Option<ChannelCongestion>)> {
         #[cfg(test)]
         {
             self.maybe_fail(crate::test_utils::FailureLocation::InsideProcessBatch)?;
@@ -73,7 +76,7 @@ impl IndexScheduler {
 
                 canceled_tasks.push(task);
 
-                Ok(canceled_tasks)
+                Ok((canceled_tasks, None))
             }
             Batch::TaskDeletions(mut tasks) => {
                 // 1. Retrieve the tasks that matched the query at enqueue-time.
@@ -112,10 +115,14 @@ impl IndexScheduler {
                         _ => unreachable!(),
                     }
                 }
-                Ok(tasks)
+                Ok((tasks, None))
+            }
+            Batch::SnapshotCreation(tasks) => {
+                self.process_snapshot(progress, tasks).map(|tasks| (tasks, None))
+            }
+            Batch::Dump(task) => {
+                self.process_dump_creation(progress, task).map(|tasks| (tasks, None))
             }
-            Batch::SnapshotCreation(tasks) => self.process_snapshot(progress, tasks),
-            Batch::Dump(task) => self.process_dump_creation(progress, task),
             Batch::IndexOperation { op, must_create_index } => {
                 let index_uid = op.index_uid().to_string();
                 let index = if must_create_index {
@@ -132,7 +139,8 @@ impl IndexScheduler {
                     .set_currently_updating_index(Some((index_uid.clone(), index.clone())));
 
                 let mut index_wtxn = index.write_txn()?;
-                let tasks = self.apply_index_operation(&mut index_wtxn, &index, op, progress)?;
+                let (tasks, congestion) =
+                    self.apply_index_operation(&mut index_wtxn, &index, op, progress)?;
 
                 {
                     let span = tracing::trace_span!(target: "indexing::scheduler", "commit");
@@ -163,7 +171,7 @@ impl IndexScheduler {
                     ),
                 }
 
-                Ok(tasks)
+                Ok((tasks, congestion))
             }
             Batch::IndexCreation { index_uid, primary_key, task } => {
                 progress.update_progress(CreateIndexProgress::CreatingTheIndex);
@@ -231,7 +239,7 @@ impl IndexScheduler {
                     ),
                 }
 
-                Ok(vec![task])
+                Ok((vec![task], None))
             }
             Batch::IndexDeletion { index_uid, index_has_been_created, mut tasks } => {
                 progress.update_progress(DeleteIndexProgress::DeletingTheIndex);
@@ -265,7 +273,7 @@ impl IndexScheduler {
                     };
                 }
 
-                Ok(tasks)
+                Ok((tasks, None))
             }
             Batch::IndexSwap { mut task } => {
                 progress.update_progress(SwappingTheIndexes::EnsuringCorrectnessOfTheSwap);
@@ -313,7 +321,7 @@ impl IndexScheduler {
                 }
                 wtxn.commit()?;
                 task.status = Status::Succeeded;
-                Ok(vec![task])
+                Ok((vec![task], None))
             }
             Batch::UpgradeDatabase { mut tasks } => {
                 let KindWithContent::UpgradeDatabase { from } = tasks.last().unwrap().kind else {
@@ -323,8 +331,17 @@ impl IndexScheduler {
                 match ret {
                     Ok(Ok(())) => (),
                     Ok(Err(e)) => return Err(Error::DatabaseUpgrade(Box::new(e))),
-                    Err(_e) => {
-                        return Err(Error::DatabaseUpgrade(Box::new(Error::ProcessBatchPanicked)));
+                    Err(e) => {
+                        let msg = match e.downcast_ref::<&'static str>() {
+                            Some(s) => *s,
+                            None => match e.downcast_ref::<String>() {
+                                Some(s) => &s[..],
+                                None => "Box<dyn Any>",
+                            },
+                        };
+                        return Err(Error::DatabaseUpgrade(Box::new(Error::ProcessBatchPanicked(
+                            msg.to_string(),
+                        ))));
                     }
                 }
 
@@ -334,7 +351,7 @@ impl IndexScheduler {
                     task.error = None;
                 }
 
-                Ok(tasks)
+                Ok((tasks, None))
             }
         }
     }
@@ -418,7 +435,6 @@ impl IndexScheduler {
         to_delete_tasks -= &enqueued_tasks;
 
         // 2. We now have a list of tasks to delete, delete them
-
         let mut affected_indexes = HashSet::new();
         let mut affected_statuses = HashSet::new();
         let mut affected_kinds = HashSet::new();
@@ -515,9 +531,51 @@ impl IndexScheduler {
                 tasks -= &to_delete_tasks;
                 // We must remove the batch entirely
                 if tasks.is_empty() {
-                    self.queue.batches.all_batches.delete(wtxn, &batch_id)?;
-                    self.queue.batch_to_tasks_mapping.delete(wtxn, &batch_id)?;
+                    if let Some(batch) = self.queue.batches.get_batch(wtxn, batch_id)? {
+                        if let Some(BatchEnqueuedAt { earliest, oldest }) = batch.enqueued_at {
+                            remove_task_datetime(
+                                wtxn,
+                                self.queue.batches.enqueued_at,
+                                earliest,
+                                batch_id,
+                            )?;
+                            remove_task_datetime(
+                                wtxn,
+                                self.queue.batches.enqueued_at,
+                                oldest,
+                                batch_id,
+                            )?;
+                        } else {
+                            // If we don't have the enqueued at in the batch it means the database comes from the v1.12
+                            // and we still need to find the date by scrolling the database
+                            remove_n_tasks_datetime_earlier_than(
+                                wtxn,
+                                self.queue.batches.enqueued_at,
+                                batch.started_at,
+                                batch.stats.total_nb_tasks.clamp(1, 2) as usize,
+                                batch_id,
+                            )?;
+                        }
+                        remove_task_datetime(
+                            wtxn,
+                            self.queue.batches.started_at,
+                            batch.started_at,
+                            batch_id,
+                        )?;
+                        if let Some(finished_at) = batch.finished_at {
+                            remove_task_datetime(
+                                wtxn,
+                                self.queue.batches.finished_at,
+                                finished_at,
+                                batch_id,
+                            )?;
+                        }
+
+                        self.queue.batches.all_batches.delete(wtxn, &batch_id)?;
+                        self.queue.batch_to_tasks_mapping.delete(wtxn, &batch_id)?;
+                    }
                 }
+
                 // Anyway, we must remove the batch from all its reverse indexes.
                 // The only way to do that is to check
 

crates/index-scheduler/src/scheduler/process_dump_creation.rs

@@ -1,3 +1,4 @@
+use std::collections::BTreeMap;
 use std::fs::File;
 use std::io::BufWriter;
 use std::sync::atomic::Ordering;
@@ -11,7 +12,9 @@ use meilisearch_types::tasks::{Details, KindWithContent, Status, Task};
 use time::macros::format_description;
 use time::OffsetDateTime;
 
-use crate::processing::{AtomicDocumentStep, AtomicTaskStep, DumpCreationProgress};
+use crate::processing::{
+    AtomicBatchStep, AtomicDocumentStep, AtomicTaskStep, DumpCreationProgress,
+};
 use crate::{Error, IndexScheduler, Result};
 
 impl IndexScheduler {
@@ -102,7 +105,40 @@ impl IndexScheduler {
         }
         dump_tasks.flush()?;
 
-        // 3. Dump the indexes
+        // 3. dump the batches
+        progress.update_progress(DumpCreationProgress::DumpTheBatches);
+        let mut dump_batches = dump.create_batches_queue()?;
+
+        let (atomic_batch_progress, update_batch_progress) =
+            AtomicBatchStep::new(self.queue.batches.all_batches.len(&rtxn)? as u32);
+        progress.update_progress(update_batch_progress);
+
+        for ret in self.queue.batches.all_batches.iter(&rtxn)? {
+            if self.scheduler.must_stop_processing.get() {
+                return Err(Error::AbortedTask);
+            }
+
+            let (_, mut b) = ret?;
+            // In the case we're dumping ourselves we want to be marked as finished
+            // to not loop over ourselves indefinitely.
+            if b.uid == task.uid {
+                let finished_at = OffsetDateTime::now_utc();
+
+                // We're going to fake the date because we don't know if everything is going to go well.
+                // But we need to dump the task as finished and successful.
+                // If something fail everything will be set appropriately in the end.
+                let mut statuses = BTreeMap::new();
+                statuses.insert(Status::Succeeded, b.stats.total_nb_tasks);
+                b.stats.status = statuses;
+                b.finished_at = Some(finished_at);
+            }
+
+            dump_batches.push_batch(&b)?;
+            atomic_batch_progress.fetch_add(1, Ordering::Relaxed);
+        }
+        dump_batches.flush()?;
+
+        // 4. Dump the indexes
         progress.update_progress(DumpCreationProgress::DumpTheIndexes);
         let nb_indexes = self.index_mapper.index_mapping.len(&rtxn)? as u32;
         let mut count = 0;
@@ -142,7 +178,7 @@ impl IndexScheduler {
             let documents = index
                 .all_documents(&rtxn)
                 .map_err(|e| Error::from_milli(e, Some(uid.to_string())))?;
-            // 3.1. Dump the documents
+            // 4.1. Dump the documents
             for ret in documents {
                 if self.scheduler.must_stop_processing.get() {
                     return Err(Error::AbortedTask);
@@ -204,7 +240,7 @@ impl IndexScheduler {
                 atomic.fetch_add(1, Ordering::Relaxed);
             }
 
-            // 3.2. Dump the settings
+            // 4.2. Dump the settings
             let settings = meilisearch_types::settings::settings(
                 index,
                 &rtxn,
@@ -215,10 +251,12 @@ impl IndexScheduler {
             Ok(())
         })?;
 
-        // 4. Dump experimental feature settings
+        // 5. Dump experimental feature settings
         progress.update_progress(DumpCreationProgress::DumpTheExperimentalFeatures);
         let features = self.features().runtime_features();
         dump.create_experimental_features(features)?;
+        let network = self.network();
+        dump.create_network(network)?;
 
         let dump_uid = started_at.format(format_description!(
                     "[year repr:full][month repr:numerical][day padding:zero]-[hour padding:zero][minute padding:zero][second padding:zero][subsecond digits:3]"

crates/index-scheduler/src/scheduler/process_index_operation.rs

@@ -5,7 +5,7 @@ use meilisearch_types::milli::documents::PrimaryKey;
 use meilisearch_types::milli::progress::Progress;
 use meilisearch_types::milli::update::new::indexer::{self, UpdateByFunction};
 use meilisearch_types::milli::update::DocumentAdditionResult;
-use meilisearch_types::milli::{self, Filter, ThreadPoolNoAbortBuilder};
+use meilisearch_types::milli::{self, ChannelCongestion, Filter, ThreadPoolNoAbortBuilder};
 use meilisearch_types::settings::apply_settings_to_builder;
 use meilisearch_types::tasks::{Details, KindWithContent, Status, Task};
 use meilisearch_types::Index;
@@ -33,9 +33,8 @@ impl IndexScheduler {
         index: &'i Index,
         operation: IndexOperation,
         progress: Progress,
-    ) -> Result<Vec<Task>> {
+    ) -> Result<(Vec<Task>, Option<ChannelCongestion>)> {
         let indexer_alloc = Bump::new();
-
         let started_processing_at = std::time::Instant::now();
         let must_stop_processing = self.scheduler.must_stop_processing.clone();
 
@@ -60,7 +59,7 @@ impl IndexScheduler {
                     };
                 }
 
-                Ok(tasks)
+                Ok((tasks, None))
             }
             IndexOperation::DocumentOperation { index_uid, primary_key, operations, mut tasks } => {
                 progress.update_progress(DocumentOperationProgress::RetrievingConfig);
@@ -173,21 +172,24 @@ impl IndexScheduler {
                 }
 
                 progress.update_progress(DocumentOperationProgress::Indexing);
+                let mut congestion = None;
                 if tasks.iter().any(|res| res.error.is_none()) {
-                    indexer::index(
-                        index_wtxn,
-                        index,
-                        pool,
-                        indexer_config.grenad_parameters(),
-                        &db_fields_ids_map,
-                        new_fields_ids_map,
-                        primary_key,
-                        &document_changes,
-                        embedders,
-                        &|| must_stop_processing.get(),
-                        &progress,
-                    )
-                    .map_err(|e| Error::from_milli(e, Some(index_uid.clone())))?;
+                    congestion = Some(
+                        indexer::index(
+                            index_wtxn,
+                            index,
+                            pool,
+                            indexer_config.grenad_parameters(),
+                            &db_fields_ids_map,
+                            new_fields_ids_map,
+                            primary_key,
+                            &document_changes,
+                            embedders,
+                            &|| must_stop_processing.get(),
+                            &progress,
+                        )
+                        .map_err(|e| Error::from_milli(e, Some(index_uid.clone())))?,
+                    );
 
                     let addition = DocumentAdditionResult {
                         indexed_documents: candidates_count,
@@ -199,7 +201,7 @@ impl IndexScheduler {
                     tracing::info!(indexing_result = ?addition, processed_in = ?started_processing_at.elapsed(), "document indexing done");
                 }
 
-                Ok(tasks)
+                Ok((tasks, congestion))
             }
             IndexOperation::DocumentEdition { index_uid, mut task } => {
                 progress.update_progress(DocumentEditionProgress::RetrievingConfig);
@@ -247,7 +249,7 @@ impl IndexScheduler {
                         edited_documents: Some(0),
                     });
 
-                    return Ok(vec![task]);
+                    return Ok((vec![task], None));
                 }
 
                 let rtxn = index.read_txn()?;
@@ -262,6 +264,7 @@ impl IndexScheduler {
 
                 let result_count = Ok((candidates.len(), candidates.len())) as Result<_>;
 
+                let mut congestion = None;
                 if task.error.is_none() {
                     let local_pool;
                     let indexer_config = self.index_mapper.indexer_config();
@@ -292,20 +295,22 @@ impl IndexScheduler {
                     let embedders = self.embedders(index_uid.clone(), embedders)?;
 
                     progress.update_progress(DocumentEditionProgress::Indexing);
-                    indexer::index(
-                        index_wtxn,
-                        index,
-                        pool,
-                        indexer_config.grenad_parameters(),
-                        &db_fields_ids_map,
-                        new_fields_ids_map,
-                        None, // cannot change primary key in DocumentEdition
-                        &document_changes,
-                        embedders,
-                        &|| must_stop_processing.get(),
-                        &progress,
-                    )
-                    .map_err(|err| Error::from_milli(err, Some(index_uid.clone())))?;
+                    congestion = Some(
+                        indexer::index(
+                            index_wtxn,
+                            index,
+                            pool,
+                            indexer_config.grenad_parameters(),
+                            &db_fields_ids_map,
+                            new_fields_ids_map,
+                            None, // cannot change primary key in DocumentEdition
+                            &document_changes,
+                            embedders,
+                            &|| must_stop_processing.get(),
+                            &progress,
+                        )
+                        .map_err(|err| Error::from_milli(err, Some(index_uid.clone())))?,
+                    );
 
                     let addition = DocumentAdditionResult {
                         indexed_documents: candidates_count,
@@ -341,7 +346,7 @@ impl IndexScheduler {
                     }
                 }
 
-                Ok(vec![task])
+                Ok((vec![task], congestion))
             }
             IndexOperation::DocumentDeletion { mut tasks, index_uid } => {
                 progress.update_progress(DocumentDeletionProgress::RetrievingConfig);
@@ -408,7 +413,7 @@ impl IndexScheduler {
                 }
 
                 if to_delete.is_empty() {
-                    return Ok(tasks);
+                    return Ok((tasks, None));
                 }
 
                 let rtxn = index.read_txn()?;
@@ -422,6 +427,7 @@ impl IndexScheduler {
                     PrimaryKey::new_or_insert(primary_key, &mut new_fields_ids_map)
                         .map_err(|err| Error::from_milli(err.into(), Some(index_uid.clone())))?;
 
+                let mut congestion = None;
                 if !tasks.iter().all(|res| res.error.is_some()) {
                     let local_pool;
                     let indexer_config = self.index_mapper.indexer_config();
@@ -447,20 +453,22 @@ impl IndexScheduler {
                     let embedders = self.embedders(index_uid.clone(), embedders)?;
 
                     progress.update_progress(DocumentDeletionProgress::Indexing);
-                    indexer::index(
-                        index_wtxn,
-                        index,
-                        pool,
-                        indexer_config.grenad_parameters(),
-                        &db_fields_ids_map,
-                        new_fields_ids_map,
-                        None, // document deletion never changes primary key
-                        &document_changes,
-                        embedders,
-                        &|| must_stop_processing.get(),
-                        &progress,
-                    )
-                    .map_err(|err| Error::from_milli(err, Some(index_uid.clone())))?;
+                    congestion = Some(
+                        indexer::index(
+                            index_wtxn,
+                            index,
+                            pool,
+                            indexer_config.grenad_parameters(),
+                            &db_fields_ids_map,
+                            new_fields_ids_map,
+                            None, // document deletion never changes primary key
+                            &document_changes,
+                            embedders,
+                            &|| must_stop_processing.get(),
+                            &progress,
+                        )
+                        .map_err(|err| Error::from_milli(err, Some(index_uid.clone())))?,
+                    );
 
                     let addition = DocumentAdditionResult {
                         indexed_documents: candidates_count,
@@ -472,7 +480,7 @@ impl IndexScheduler {
                     tracing::info!(indexing_result = ?addition, processed_in = ?started_processing_at.elapsed(), "document indexing done");
                 }
 
-                Ok(tasks)
+                Ok((tasks, congestion))
             }
             IndexOperation::Settings { index_uid, settings, mut tasks } => {
                 progress.update_progress(SettingsProgress::RetrievingAndMergingTheSettings);
@@ -497,7 +505,7 @@ impl IndexScheduler {
                     )
                     .map_err(|err| Error::from_milli(err, Some(index_uid.clone())))?;
 
-                Ok(tasks)
+                Ok((tasks, None))
             }
             IndexOperation::DocumentClearAndSetting {
                 index_uid,
@@ -505,7 +513,7 @@ impl IndexScheduler {
                 settings,
                 settings_tasks,
             } => {
-                let mut import_tasks = self.apply_index_operation(
+                let (mut import_tasks, _congestion) = self.apply_index_operation(
                     index_wtxn,
                     index,
                     IndexOperation::DocumentClear {
@@ -515,7 +523,7 @@ impl IndexScheduler {
                     progress.clone(),
                 )?;
 
-                let settings_tasks = self.apply_index_operation(
+                let (settings_tasks, _congestion) = self.apply_index_operation(
                     index_wtxn,
                     index,
                     IndexOperation::Settings { index_uid, settings, tasks: settings_tasks },
@@ -524,7 +532,7 @@ impl IndexScheduler {
 
                 let mut tasks = settings_tasks;
                 tasks.append(&mut import_tasks);
-                Ok(tasks)
+                Ok((tasks, None))
             }
         }
     }

crates/index-scheduler/src/scheduler/snapshots/index_scheduler__scheduler__test_embedders__import_vectors-8.snap

@@ -1,12 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test_embedders.rs
 expression: simple_hf_config.embedder_options
-snapshot_kind: text
 ---
 {
   "HuggingFace": {
     "model": "sentence-transformers/all-MiniLM-L6-v2",
     "revision": "e4ce9877abf3edfe10b0d82785e83bdcb973e22e",
-    "distribution": null
+    "distribution": null,
+    "pooling": "useModel"
   }
 }

crates/index-scheduler/src/scheduler/snapshots/test.rs/task_deletion_delete_same_task_twice/task_deletion_processed.snap

@@ -56,16 +56,13 @@ succeeded [1,]
 ### Batches Index Tasks:
 ----------------------------------------------------------------------
 ### Batches Enqueued At:
-[timestamp] [0,]
 [timestamp] [1,]
 [timestamp] [1,]
 ----------------------------------------------------------------------
 ### Batches Started At:
-[timestamp] [0,]
 [timestamp] [1,]
 ----------------------------------------------------------------------
 ### Batches Finished At:
-[timestamp] [0,]
 [timestamp] [1,]
 ----------------------------------------------------------------------
 ### File Store:

crates/index-scheduler/src/scheduler/snapshots/test.rs/task_deletion_deleteable/task_deletion_processed.snap

@@ -54,15 +54,12 @@ succeeded [1,]
 ### Batches Index Tasks:
 ----------------------------------------------------------------------
 ### Batches Enqueued At:
-[timestamp] [0,]
 [timestamp] [1,]
 ----------------------------------------------------------------------
 ### Batches Started At:
-[timestamp] [0,]
 [timestamp] [1,]
 ----------------------------------------------------------------------
 ### Batches Finished At:
-[timestamp] [0,]
 [timestamp] [1,]
 ----------------------------------------------------------------------
 ### File Store:

crates/index-scheduler/src/scheduler/snapshots/test.rs/test_settings_update/after_registering_settings_task.snap

@@ -1,13 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test.rs
-snapshot_kind: text
 ---
 ### Autobatching Enabled = true
 ### Processing batch None:
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, status: enqueued, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"default": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, api_key: Set("My super secret"), dimensions: Set(4), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"default": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, api_key: Set("My super secret"), dimensions: Set(4), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
+0 {uid: 0, status: enqueued, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"default": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, pooling: NotSet, api_key: Set("My super secret"), dimensions: Set(4), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, search_embedder: NotSet, indexing_embedder: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"default": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, pooling: NotSet, api_key: Set("My super secret"), dimensions: Set(4), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, search_embedder: NotSet, indexing_embedder: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
 ----------------------------------------------------------------------
 ### Status:
 enqueued [0,]

crates/index-scheduler/src/scheduler/snapshots/test.rs/test_settings_update/settings_update_processed.snap

@@ -1,13 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test.rs
-snapshot_kind: text
 ---
 ### Autobatching Enabled = true
 ### Processing batch None:
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, batch_uid: 0, status: succeeded, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"default": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, api_key: Set("My super secret"), dimensions: Set(4), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"default": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, api_key: Set("My super secret"), dimensions: Set(4), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
+0 {uid: 0, batch_uid: 0, status: succeeded, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"default": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, pooling: NotSet, api_key: Set("My super secret"), dimensions: Set(4), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, search_embedder: NotSet, indexing_embedder: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"default": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, pooling: NotSet, api_key: Set("My super secret"), dimensions: Set(4), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, search_embedder: NotSet, indexing_embedder: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
 ----------------------------------------------------------------------
 ### Status:
 enqueued []

crates/index-scheduler/src/scheduler/snapshots/test_embedders.rs/import_vectors/Intel to kefir succeeds.snap

@@ -1,13 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test_embedders.rs
-snapshot_kind: text
 ---
 ### Autobatching Enabled = true
 ### Processing batch None:
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, batch_uid: 0, status: succeeded, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
+0 {uid: 0, batch_uid: 0, status: succeeded, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, pooling: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, search_embedder: NotSet, indexing_embedder: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), pooling: NotSet, api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, search_embedder: NotSet, indexing_embedder: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, pooling: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, search_embedder: NotSet, indexing_embedder: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), pooling: NotSet, api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, search_embedder: NotSet, indexing_embedder: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
 1 {uid: 1, batch_uid: 1, status: succeeded, details: { received_documents: 1, indexed_documents: Some(1) }, kind: DocumentAdditionOrUpdate { index_uid: "doggos", primary_key: Some("id"), method: UpdateDocuments, content_file: 00000000-0000-0000-0000-000000000000, documents_count: 1, allow_index_creation: true }}
 2 {uid: 2, batch_uid: 2, status: succeeded, details: { received_documents: 1, indexed_documents: Some(1) }, kind: DocumentAdditionOrUpdate { index_uid: "doggos", primary_key: None, method: UpdateDocuments, content_file: 00000000-0000-0000-0000-000000000001, documents_count: 1, allow_index_creation: true }}
 ----------------------------------------------------------------------

crates/index-scheduler/src/scheduler/snapshots/test_embedders.rs/import_vectors/Intel to kefir.snap

@@ -1,13 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test_embedders.rs
-snapshot_kind: text
 ---
 ### Autobatching Enabled = true
 ### Processing batch None:
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, batch_uid: 0, status: succeeded, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
+0 {uid: 0, batch_uid: 0, status: succeeded, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, pooling: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, search_embedder: NotSet, indexing_embedder: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), pooling: NotSet, api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, search_embedder: NotSet, indexing_embedder: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, pooling: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, search_embedder: NotSet, indexing_embedder: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), pooling: NotSet, api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, search_embedder: NotSet, indexing_embedder: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
 1 {uid: 1, batch_uid: 1, status: succeeded, details: { received_documents: 1, indexed_documents: Some(1) }, kind: DocumentAdditionOrUpdate { index_uid: "doggos", primary_key: Some("id"), method: UpdateDocuments, content_file: 00000000-0000-0000-0000-000000000000, documents_count: 1, allow_index_creation: true }}
 2 {uid: 2, status: enqueued, details: { received_documents: 1, indexed_documents: None }, kind: DocumentAdditionOrUpdate { index_uid: "doggos", primary_key: None, method: UpdateDocuments, content_file: 00000000-0000-0000-0000-000000000001, documents_count: 1, allow_index_creation: true }}
 ----------------------------------------------------------------------

crates/index-scheduler/src/scheduler/snapshots/test_embedders.rs/import_vectors/adding Intel succeeds.snap

@@ -1,13 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test_embedders.rs
-snapshot_kind: text
 ---
 ### Autobatching Enabled = true
 ### Processing batch None:
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, batch_uid: 0, status: succeeded, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
+0 {uid: 0, batch_uid: 0, status: succeeded, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, pooling: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, search_embedder: NotSet, indexing_embedder: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), pooling: NotSet, api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, search_embedder: NotSet, indexing_embedder: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, pooling: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, search_embedder: NotSet, indexing_embedder: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), pooling: NotSet, api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, search_embedder: NotSet, indexing_embedder: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
 1 {uid: 1, batch_uid: 1, status: succeeded, details: { received_documents: 1, indexed_documents: Some(1) }, kind: DocumentAdditionOrUpdate { index_uid: "doggos", primary_key: Some("id"), method: UpdateDocuments, content_file: 00000000-0000-0000-0000-000000000000, documents_count: 1, allow_index_creation: true }}
 ----------------------------------------------------------------------
 ### Status:

crates/index-scheduler/src/scheduler/snapshots/test_embedders.rs/import_vectors/after adding Intel.snap

@@ -1,13 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test_embedders.rs
-snapshot_kind: text
 ---
 ### Autobatching Enabled = true
 ### Processing batch None:
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, batch_uid: 0, status: succeeded, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
+0 {uid: 0, batch_uid: 0, status: succeeded, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, pooling: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, search_embedder: NotSet, indexing_embedder: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), pooling: NotSet, api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, search_embedder: NotSet, indexing_embedder: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, pooling: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, search_embedder: NotSet, indexing_embedder: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), pooling: NotSet, api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, search_embedder: NotSet, indexing_embedder: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
 1 {uid: 1, status: enqueued, details: { received_documents: 1, indexed_documents: None }, kind: DocumentAdditionOrUpdate { index_uid: "doggos", primary_key: Some("id"), method: UpdateDocuments, content_file: 00000000-0000-0000-0000-000000000000, documents_count: 1, allow_index_creation: true }}
 ----------------------------------------------------------------------
 ### Status:

crates/index-scheduler/src/scheduler/snapshots/test_embedders.rs/import_vectors/after_registering_settings_task_vectors.snap

@@ -1,13 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test_embedders.rs
-snapshot_kind: text
 ---
 ### Autobatching Enabled = true
 ### Processing batch None:
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, status: enqueued, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
+0 {uid: 0, status: enqueued, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, pooling: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, search_embedder: NotSet, indexing_embedder: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), pooling: NotSet, api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, search_embedder: NotSet, indexing_embedder: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, pooling: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, search_embedder: NotSet, indexing_embedder: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), pooling: NotSet, api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, search_embedder: NotSet, indexing_embedder: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
 ----------------------------------------------------------------------
 ### Status:
 enqueued [0,]

crates/index-scheduler/src/scheduler/snapshots/test_embedders.rs/import_vectors/settings_update_processed_vectors.snap

@@ -1,13 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test_embedders.rs
-snapshot_kind: text
 ---
 ### Autobatching Enabled = true
 ### Processing batch None:
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, batch_uid: 0, status: succeeded, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
+0 {uid: 0, batch_uid: 0, status: succeeded, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, pooling: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, search_embedder: NotSet, indexing_embedder: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), pooling: NotSet, api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, search_embedder: NotSet, indexing_embedder: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: NotSet, sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: Set({"A_fakerest": Set(EmbeddingSettings { source: Set(Rest), model: NotSet, revision: NotSet, pooling: NotSet, api_key: Set("My super secret"), dimensions: Set(384), binary_quantized: NotSet, document_template: NotSet, document_template_max_bytes: NotSet, url: Set("http://localhost:7777"), request: Set(String("{{text}}")), response: Set(String("{{embedding}}")), headers: NotSet, search_embedder: NotSet, indexing_embedder: NotSet, distribution: NotSet }), "B_small_hf": Set(EmbeddingSettings { source: Set(HuggingFace), model: Set("sentence-transformers/all-MiniLM-L6-v2"), revision: Set("e4ce9877abf3edfe10b0d82785e83bdcb973e22e"), pooling: NotSet, api_key: NotSet, dimensions: NotSet, binary_quantized: NotSet, document_template: Set("{{doc.doggo}} the {{doc.breed}} best doggo"), document_template_max_bytes: NotSet, url: NotSet, request: NotSet, response: NotSet, headers: NotSet, search_embedder: NotSet, indexing_embedder: NotSet, distribution: NotSet })}), search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
 ----------------------------------------------------------------------
 ### Status:
 enqueued []

crates/index-scheduler/src/scheduler/snapshots/test_failure.rs/fail_in_process_batch_for_document_deletion/after_adding_the_documents.snap

@@ -1,13 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test_failure.rs
-snapshot_kind: text
 ---
 ### Autobatching Enabled = true
 ### Processing batch None:
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, batch_uid: 0, status: succeeded, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: Set({"catto"}), sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: NotSet, search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: Set({"catto"}), sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: NotSet, search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
+0 {uid: 0, batch_uid: 0, status: succeeded, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: Set([Field("catto")]), sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: NotSet, search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: Set([Field("catto")]), sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: NotSet, search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
 1 {uid: 1, batch_uid: 1, status: succeeded, details: { received_documents: 3, indexed_documents: Some(3) }, kind: DocumentAdditionOrUpdate { index_uid: "doggos", primary_key: Some("id"), method: ReplaceDocuments, content_file: 00000000-0000-0000-0000-000000000000, documents_count: 3, allow_index_creation: true }}
 ----------------------------------------------------------------------
 ### Status:

crates/index-scheduler/src/scheduler/snapshots/test_failure.rs/fail_in_process_batch_for_document_deletion/after_adding_the_settings.snap

@@ -1,13 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test_failure.rs
-snapshot_kind: text
 ---
 ### Autobatching Enabled = true
 ### Processing batch None:
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, batch_uid: 0, status: succeeded, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: Set({"catto"}), sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: NotSet, search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: Set({"catto"}), sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: NotSet, search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
+0 {uid: 0, batch_uid: 0, status: succeeded, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: Set([Field("catto")]), sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: NotSet, search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: Set([Field("catto")]), sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: NotSet, search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
 1 {uid: 1, status: enqueued, details: { received_documents: 3, indexed_documents: None }, kind: DocumentAdditionOrUpdate { index_uid: "doggos", primary_key: Some("id"), method: ReplaceDocuments, content_file: 00000000-0000-0000-0000-000000000000, documents_count: 3, allow_index_creation: true }}
 ----------------------------------------------------------------------
 ### Status:

crates/index-scheduler/src/scheduler/snapshots/test_failure.rs/fail_in_process_batch_for_document_deletion/after_removing_the_documents.snap

@@ -1,17 +1,16 @@
 ---
 source: crates/index-scheduler/src/scheduler/test_failure.rs
-snapshot_kind: text
 ---
 ### Autobatching Enabled = true
 ### Processing batch None:
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, batch_uid: 0, status: succeeded, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: Set({"catto"}), sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: NotSet, search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: Set({"catto"}), sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: NotSet, search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
+0 {uid: 0, batch_uid: 0, status: succeeded, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: Set([Field("catto")]), sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: NotSet, search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: Set([Field("catto")]), sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: NotSet, search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
 1 {uid: 1, batch_uid: 1, status: succeeded, details: { received_documents: 3, indexed_documents: Some(3) }, kind: DocumentAdditionOrUpdate { index_uid: "doggos", primary_key: Some("id"), method: ReplaceDocuments, content_file: 00000000-0000-0000-0000-000000000000, documents_count: 3, allow_index_creation: true }}
 2 {uid: 2, batch_uid: 2, status: succeeded, details: { received_document_ids: 1, deleted_documents: Some(1) }, kind: DocumentDeletion { index_uid: "doggos", documents_ids: ["1"] }}
 3 {uid: 3, batch_uid: 2, status: failed, error: ResponseError { code: 200, message: "Index `doggos`: Invalid type for filter subexpression: expected: String, Array, found: true.", error_code: "invalid_document_filter", error_type: "invalid_request", error_link: "https://docs.meilisearch.com/errors#invalid_document_filter" }, details: { original_filter: true, deleted_documents: Some(0) }, kind: DocumentDeletionByFilter { index_uid: "doggos", filter_expr: Bool(true) }}
-4 {uid: 4, batch_uid: 2, status: failed, error: ResponseError { code: 200, message: "Index `doggos`: Attribute `id` is not filterable. Available filterable attributes are: `catto`.\n1:3 id = 2", error_code: "invalid_document_filter", error_type: "invalid_request", error_link: "https://docs.meilisearch.com/errors#invalid_document_filter" }, details: { original_filter: "id = 2", deleted_documents: Some(0) }, kind: DocumentDeletionByFilter { index_uid: "doggos", filter_expr: String("id = 2") }}
+4 {uid: 4, batch_uid: 2, status: failed, error: ResponseError { code: 200, message: "Index `doggos`: Attribute `id` is not filterable. Available filterable attribute patterns are: `catto`.\n1:3 id = 2", error_code: "invalid_document_filter", error_type: "invalid_request", error_link: "https://docs.meilisearch.com/errors#invalid_document_filter" }, details: { original_filter: "id = 2", deleted_documents: Some(0) }, kind: DocumentDeletionByFilter { index_uid: "doggos", filter_expr: String("id = 2") }}
 5 {uid: 5, batch_uid: 2, status: succeeded, details: { original_filter: "catto EXISTS", deleted_documents: Some(1) }, kind: DocumentDeletionByFilter { index_uid: "doggos", filter_expr: String("catto EXISTS") }}
 ----------------------------------------------------------------------
 ### Status:

crates/index-scheduler/src/scheduler/snapshots/test_failure.rs/fail_in_process_batch_for_document_deletion/registered_the_document_deletions.snap

@@ -1,13 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test_failure.rs
-snapshot_kind: text
 ---
 ### Autobatching Enabled = true
 ### Processing batch None:
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, batch_uid: 0, status: succeeded, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: Set({"catto"}), sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: NotSet, search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: Set({"catto"}), sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: NotSet, search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
+0 {uid: 0, batch_uid: 0, status: succeeded, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: Set([Field("catto")]), sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: NotSet, search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: Set([Field("catto")]), sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: NotSet, search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
 1 {uid: 1, batch_uid: 1, status: succeeded, details: { received_documents: 3, indexed_documents: Some(3) }, kind: DocumentAdditionOrUpdate { index_uid: "doggos", primary_key: Some("id"), method: ReplaceDocuments, content_file: 00000000-0000-0000-0000-000000000000, documents_count: 3, allow_index_creation: true }}
 2 {uid: 2, status: enqueued, details: { received_document_ids: 1, deleted_documents: None }, kind: DocumentDeletion { index_uid: "doggos", documents_ids: ["1"] }}
 3 {uid: 3, status: enqueued, details: { original_filter: true, deleted_documents: None }, kind: DocumentDeletionByFilter { index_uid: "doggos", filter_expr: Bool(true) }}

crates/index-scheduler/src/scheduler/snapshots/test_failure.rs/fail_in_process_batch_for_document_deletion/registered_the_setting_and_document_addition.snap

@@ -1,13 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test_failure.rs
-snapshot_kind: text
 ---
 ### Autobatching Enabled = true
 ### Processing batch None:
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, status: enqueued, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: Set({"catto"}), sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: NotSet, search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: Set({"catto"}), sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: NotSet, search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
+0 {uid: 0, status: enqueued, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: Set([Field("catto")]), sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: NotSet, search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: Set([Field("catto")]), sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: NotSet, search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
 1 {uid: 1, status: enqueued, details: { received_documents: 3, indexed_documents: None }, kind: DocumentAdditionOrUpdate { index_uid: "doggos", primary_key: Some("id"), method: ReplaceDocuments, content_file: 00000000-0000-0000-0000-000000000000, documents_count: 3, allow_index_creation: true }}
 ----------------------------------------------------------------------
 ### Status:

crates/index-scheduler/src/scheduler/snapshots/test_failure.rs/panic_in_process_batch_for_index_creation/index_creation_failed.snap

@@ -7,7 +7,7 @@ snapshot_kind: text
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, batch_uid: 0, status: failed, error: ResponseError { code: 200, message: "An unexpected crash occurred when processing the task.", error_code: "internal", error_type: "internal", error_link: "https://docs.meilisearch.com/errors#internal" }, details: { primary_key: Some("mouse") }, kind: IndexCreation { index_uid: "catto", primary_key: Some("mouse") }}
+0 {uid: 0, batch_uid: 0, status: failed, error: ResponseError { code: 200, message: "An unexpected crash occurred when processing the task: simulated panic", error_code: "internal", error_type: "internal", error_link: "https://docs.meilisearch.com/errors#internal" }, details: { primary_key: Some("mouse") }, kind: IndexCreation { index_uid: "catto", primary_key: Some("mouse") }}
 ----------------------------------------------------------------------
 ### Status:
 enqueued []

crates/index-scheduler/src/scheduler/snapshots/test_failure.rs/upgrade_failure/after_processing_everything.snap

@@ -1,13 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test_failure.rs
-snapshot_kind: text
 ---
 ### Autobatching Enabled = true
 ### Processing batch None:
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, batch_uid: 0, status: succeeded, details: { from: (1, 12, 0), to: (1, 13, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
+0 {uid: 0, batch_uid: 0, status: succeeded, details: { from: (1, 12, 0), to: (1, 13, 3) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
 1 {uid: 1, batch_uid: 1, status: succeeded, details: { primary_key: Some("mouse") }, kind: IndexCreation { index_uid: "catto", primary_key: Some("mouse") }}
 2 {uid: 2, batch_uid: 2, status: succeeded, details: { primary_key: Some("bone") }, kind: IndexCreation { index_uid: "doggo", primary_key: Some("bone") }}
 3 {uid: 3, batch_uid: 3, status: failed, error: ResponseError { code: 200, message: "Index `doggo` already exists.", error_code: "index_already_exists", error_type: "invalid_request", error_link: "https://docs.meilisearch.com/errors#index_already_exists" }, details: { primary_key: Some("bone") }, kind: IndexCreation { index_uid: "doggo", primary_key: Some("bone") }}
@@ -58,7 +57,7 @@ girafo: { number_of_documents: 0, field_distribution: {} }
 [timestamp] [4,]
 ----------------------------------------------------------------------
 ### All Batches:
-0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.13.0"}, stats: {"totalNbTasks":1,"status":{"succeeded":1},"types":{"upgradeDatabase":1},"indexUids":{}}, }
+0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.13.3"}, stats: {"totalNbTasks":1,"status":{"succeeded":1},"types":{"upgradeDatabase":1},"indexUids":{}}, }
 1 {uid: 1, details: {"primaryKey":"mouse"}, stats: {"totalNbTasks":1,"status":{"succeeded":1},"types":{"indexCreation":1},"indexUids":{"catto":1}}, }
 2 {uid: 2, details: {"primaryKey":"bone"}, stats: {"totalNbTasks":1,"status":{"succeeded":1},"types":{"indexCreation":1},"indexUids":{"doggo":1}}, }
 3 {uid: 3, details: {"primaryKey":"bone"}, stats: {"totalNbTasks":1,"status":{"failed":1},"types":{"indexCreation":1},"indexUids":{"doggo":1}}, }

crates/index-scheduler/src/scheduler/snapshots/test_failure.rs/upgrade_failure/after_removing_the_upgrade_tasks.snap

@@ -87,7 +87,6 @@ doggo [2,3,]
 girafo [4,]
 ----------------------------------------------------------------------
 ### Batches Enqueued At:
-[timestamp] [0,]
 [timestamp] [1,]
 [timestamp] [2,]
 [timestamp] [3,]
@@ -95,7 +94,6 @@ girafo [4,]
 [timestamp] [5,]
 ----------------------------------------------------------------------
 ### Batches Started At:
-[timestamp] [0,]
 [timestamp] [1,]
 [timestamp] [2,]
 [timestamp] [3,]
@@ -103,7 +101,6 @@ girafo [4,]
 [timestamp] [5,]
 ----------------------------------------------------------------------
 ### Batches Finished At:
-[timestamp] [0,]
 [timestamp] [1,]
 [timestamp] [2,]
 [timestamp] [3,]

crates/index-scheduler/src/scheduler/snapshots/test_failure.rs/upgrade_failure/register_automatic_upgrade_task.snap

@@ -1,13 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test_failure.rs
-snapshot_kind: text
 ---
 ### Autobatching Enabled = true
 ### Processing batch None:
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, status: enqueued, details: { from: (1, 12, 0), to: (1, 13, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
+0 {uid: 0, status: enqueued, details: { from: (1, 12, 0), to: (1, 13, 3) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
 ----------------------------------------------------------------------
 ### Status:
 enqueued [0,]

crates/index-scheduler/src/scheduler/snapshots/test_failure.rs/upgrade_failure/registered_a_task_while_the_upgrade_task_is_enqueued.snap

@@ -1,13 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test_failure.rs
-snapshot_kind: text
 ---
 ### Autobatching Enabled = true
 ### Processing batch None:
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, status: enqueued, details: { from: (1, 12, 0), to: (1, 13, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
+0 {uid: 0, status: enqueued, details: { from: (1, 12, 0), to: (1, 13, 3) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
 1 {uid: 1, status: enqueued, details: { primary_key: Some("mouse") }, kind: IndexCreation { index_uid: "catto", primary_key: Some("mouse") }}
 ----------------------------------------------------------------------
 ### Status:

crates/index-scheduler/src/scheduler/snapshots/test_failure.rs/upgrade_failure/upgrade_task_failed.snap

@@ -1,13 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test_failure.rs
-snapshot_kind: text
 ---
 ### Autobatching Enabled = true
 ### Processing batch None:
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, batch_uid: 0, status: failed, error: ResponseError { code: 200, message: "Planned failure for tests.", error_code: "internal", error_type: "internal", error_link: "https://docs.meilisearch.com/errors#internal" }, details: { from: (1, 12, 0), to: (1, 13, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
+0 {uid: 0, batch_uid: 0, status: failed, error: ResponseError { code: 200, message: "Planned failure for tests.", error_code: "internal", error_type: "internal", error_link: "https://docs.meilisearch.com/errors#internal" }, details: { from: (1, 12, 0), to: (1, 13, 3) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
 1 {uid: 1, status: enqueued, details: { primary_key: Some("mouse") }, kind: IndexCreation { index_uid: "catto", primary_key: Some("mouse") }}
 ----------------------------------------------------------------------
 ### Status:
@@ -38,7 +37,7 @@ catto [1,]
 [timestamp] [0,]
 ----------------------------------------------------------------------
 ### All Batches:
-0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.13.0"}, stats: {"totalNbTasks":1,"status":{"failed":1},"types":{"upgradeDatabase":1},"indexUids":{}}, }
+0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.13.3"}, stats: {"totalNbTasks":1,"status":{"failed":1},"types":{"upgradeDatabase":1},"indexUids":{}}, }
 ----------------------------------------------------------------------
 ### Batch to tasks mapping:
 0 [0,]

crates/index-scheduler/src/scheduler/snapshots/test_failure.rs/upgrade_failure/upgrade_task_failed_again.snap

@@ -1,13 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test_failure.rs
-snapshot_kind: text
 ---
 ### Autobatching Enabled = true
 ### Processing batch None:
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, batch_uid: 0, status: failed, error: ResponseError { code: 200, message: "Planned failure for tests.", error_code: "internal", error_type: "internal", error_link: "https://docs.meilisearch.com/errors#internal" }, details: { from: (1, 12, 0), to: (1, 13, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
+0 {uid: 0, batch_uid: 0, status: failed, error: ResponseError { code: 200, message: "Planned failure for tests.", error_code: "internal", error_type: "internal", error_link: "https://docs.meilisearch.com/errors#internal" }, details: { from: (1, 12, 0), to: (1, 13, 3) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
 1 {uid: 1, status: enqueued, details: { primary_key: Some("mouse") }, kind: IndexCreation { index_uid: "catto", primary_key: Some("mouse") }}
 2 {uid: 2, status: enqueued, details: { primary_key: Some("bone") }, kind: IndexCreation { index_uid: "doggo", primary_key: Some("bone") }}
 ----------------------------------------------------------------------
@@ -41,7 +40,7 @@ doggo [2,]
 [timestamp] [0,]
 ----------------------------------------------------------------------
 ### All Batches:
-0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.13.0"}, stats: {"totalNbTasks":1,"status":{"failed":1},"types":{"upgradeDatabase":1},"indexUids":{}}, }
+0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.13.3"}, stats: {"totalNbTasks":1,"status":{"failed":1},"types":{"upgradeDatabase":1},"indexUids":{}}, }
 ----------------------------------------------------------------------
 ### Batch to tasks mapping:
 0 [0,]

crates/index-scheduler/src/scheduler/snapshots/test_failure.rs/upgrade_failure/upgrade_task_succeeded.snap

@@ -1,13 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test_failure.rs
-snapshot_kind: text
 ---
 ### Autobatching Enabled = true
 ### Processing batch None:
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, batch_uid: 0, status: succeeded, details: { from: (1, 12, 0), to: (1, 13, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
+0 {uid: 0, batch_uid: 0, status: succeeded, details: { from: (1, 12, 0), to: (1, 13, 3) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
 1 {uid: 1, status: enqueued, details: { primary_key: Some("mouse") }, kind: IndexCreation { index_uid: "catto", primary_key: Some("mouse") }}
 2 {uid: 2, status: enqueued, details: { primary_key: Some("bone") }, kind: IndexCreation { index_uid: "doggo", primary_key: Some("bone") }}
 3 {uid: 3, status: enqueued, details: { primary_key: Some("bone") }, kind: IndexCreation { index_uid: "doggo", primary_key: Some("bone") }}
@@ -44,7 +43,7 @@ doggo [2,3,]
 [timestamp] [0,]
 ----------------------------------------------------------------------
 ### All Batches:
-0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.13.0"}, stats: {"totalNbTasks":1,"status":{"succeeded":1},"types":{"upgradeDatabase":1},"indexUids":{}}, }
+0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.13.3"}, stats: {"totalNbTasks":1,"status":{"succeeded":1},"types":{"upgradeDatabase":1},"indexUids":{}}, }
 ----------------------------------------------------------------------
 ### Batch to tasks mapping:
 0 [0,]

crates/index-scheduler/src/scheduler/test.rs

@@ -903,15 +903,21 @@ fn create_and_list_index() {
 
     index_scheduler.index("kefir").unwrap();
     let list = index_scheduler.get_paginated_indexes_stats(&AuthFilter::default(), 0, 20).unwrap();
-    snapshot!(json_string!(list, { "[1][0][1].created_at" => "[date]", "[1][0][1].updated_at" => "[date]", "[1][0][1].used_database_size" => "[bytes]", "[1][0][1].database_size" => "[bytes]" }), @r#"
+    snapshot!(json_string!(list, { "[1][0][1].created_at" => "[date]", "[1][0][1].updated_at" => "[date]", "[1][0][1].used_database_size" => "[bytes]", "[1][0][1].database_size" => "[bytes]" }), @r###"
     [
       1,
       [
         [
           "kefir",
           {
-            "number_of_documents": 0,
+            "documents_database_stats": {
+              "numberOfEntries": 0,
+              "totalKeySize": 0,
+              "totalValueSize": 0
+            },
             "database_size": "[bytes]",
+            "number_of_embeddings": 0,
+            "number_of_embedded_documents": 0,
             "used_database_size": "[bytes]",
             "primary_key": null,
             "field_distribution": {},
@@ -921,5 +927,5 @@ fn create_and_list_index() {
         ]
       ]
     ]
-    "#);
+    "###);
 }

crates/index-scheduler/src/scheduler/test_embedders.rs

@@ -104,9 +104,10 @@ fn import_vectors() {
 
         let configs = index_scheduler.embedders("doggos".to_string(), configs).unwrap();
         let (hf_embedder, _, _) = configs.get(&simple_hf_name).unwrap();
-        let beagle_embed = hf_embedder.embed_one(S("Intel the beagle best doggo"), None).unwrap();
-        let lab_embed = hf_embedder.embed_one(S("Max the lab best doggo"), None).unwrap();
-        let patou_embed = hf_embedder.embed_one(S("kefir the patou best doggo"), None).unwrap();
+        let beagle_embed =
+            hf_embedder.embed_search(S("Intel the beagle best doggo"), None).unwrap();
+        let lab_embed = hf_embedder.embed_search(S("Max the lab best doggo"), None).unwrap();
+        let patou_embed = hf_embedder.embed_search(S("kefir the patou best doggo"), None).unwrap();
         (fakerest_name, simple_hf_name, beagle_embed, lab_embed, patou_embed)
     };
 
@@ -404,31 +405,32 @@ fn import_vectors_first_and_embedder_later() {
     // even though we specified the vector for the ID 3, it shouldn't be marked
     // as user provided since we explicitely marked it as NOT user provided.
     snapshot!(format!("{conf:#?}"), @r###"
-        [
-            IndexEmbeddingConfig {
-                name: "my_doggo_embedder",
-                config: EmbeddingConfig {
-                    embedder_options: HuggingFace(
-                        EmbedderOptions {
-                            model: "sentence-transformers/all-MiniLM-L6-v2",
-                            revision: Some(
-                                "e4ce9877abf3edfe10b0d82785e83bdcb973e22e",
-                            ),
-                            distribution: None,
-                        },
-                    ),
-                    prompt: PromptData {
-                        template: "{{doc.doggo}}",
-                        max_bytes: Some(
-                            400,
+    [
+        IndexEmbeddingConfig {
+            name: "my_doggo_embedder",
+            config: EmbeddingConfig {
+                embedder_options: HuggingFace(
+                    EmbedderOptions {
+                        model: "sentence-transformers/all-MiniLM-L6-v2",
+                        revision: Some(
+                            "e4ce9877abf3edfe10b0d82785e83bdcb973e22e",
                         ),
+                        distribution: None,
+                        pooling: UseModel,
                     },
-                    quantized: None,
+                ),
+                prompt: PromptData {
+                    template: "{{doc.doggo}}",
+                    max_bytes: Some(
+                        400,
+                    ),
                 },
-                user_provided: RoaringBitmap<[1, 2]>,
+                quantized: None,
             },
-        ]
-        "###);
+            user_provided: RoaringBitmap<[1, 2]>,
+        },
+    ]
+    "###);
     let docid = index.external_documents_ids.get(&rtxn, "0").unwrap().unwrap();
     let embeddings = index.embeddings(&rtxn, docid).unwrap();
     let embedding = &embeddings["my_doggo_embedder"];

crates/index-scheduler/src/scheduler/test_failure.rs

@@ -1,13 +1,12 @@
 use std::time::Instant;
 
 use big_s::S;
-use maplit::btreeset;
 use meili_snap::snapshot;
 use meilisearch_types::milli::obkv_to_json;
 use meilisearch_types::milli::update::IndexDocumentsMethod::*;
 use meilisearch_types::milli::update::Setting;
-use meilisearch_types::tasks::Kind;
-use meilisearch_types::tasks::KindWithContent;
+use meilisearch_types::milli::FilterableAttributesRule;
+use meilisearch_types::tasks::{Kind, KindWithContent};
 
 use crate::insta_snapshot::snapshot_index_scheduler;
 use crate::test_utils::Breakpoint::*;
@@ -128,7 +127,8 @@ fn fail_in_process_batch_for_document_deletion() {
 
     use meilisearch_types::settings::{Settings, Unchecked};
     let mut new_settings: Box<Settings<Unchecked>> = Box::default();
-    new_settings.filterable_attributes = Setting::Set(btreeset!(S("catto")));
+    new_settings.filterable_attributes =
+        Setting::Set(vec![FilterableAttributesRule::Field(S("catto"))]);
 
     index_scheduler
         .register(

crates/index-scheduler/src/upgrade/mod.rs

@@ -24,10 +24,11 @@ pub fn upgrade_index_scheduler(
     let current_minor = to.1;
     let current_patch = to.2;
 
-    let upgrade_functions: &[&dyn UpgradeIndexScheduler] = &[&V1_12_ToCurrent {}];
+    let upgrade_functions: &[&dyn UpgradeIndexScheduler] = &[&ToCurrentNoOp {}];
 
     let start = match from {
         (1, 12, _) => 0,
+        (1, 13, _) => 0,
         (major, minor, patch) => {
             if major > current_major
                 || (major == current_major && minor > current_minor)
@@ -46,20 +47,19 @@ pub fn upgrade_index_scheduler(
         }
     };
 
-    let mut current_version = from;
-
     info!("Upgrading the task queue");
+    let mut local_from = from;
     for upgrade in upgrade_functions[start..].iter() {
         let target = upgrade.target_version();
         info!(
             "Upgrading from v{}.{}.{} to v{}.{}.{}",
-            from.0, from.1, from.2, current_version.0, current_version.1, current_version.2
+            local_from.0, local_from.1, local_from.2, target.0, target.1, target.2
         );
         let mut wtxn = env.write_txn()?;
-        upgrade.upgrade(env, &mut wtxn, from)?;
+        upgrade.upgrade(env, &mut wtxn, local_from)?;
         versioning.set_version(&mut wtxn, target)?;
         wtxn.commit()?;
-        current_version = target;
+        local_from = target;
     }
 
     let mut wtxn = env.write_txn()?;
@@ -86,9 +86,9 @@ pub fn upgrade_index_scheduler(
 }
 
 #[allow(non_camel_case_types)]
-struct V1_12_ToCurrent {}
+struct ToCurrentNoOp {}
 
-impl UpgradeIndexScheduler for V1_12_ToCurrent {
+impl UpgradeIndexScheduler for ToCurrentNoOp {
     fn upgrade(
         &self,
         _env: &Env,

crates/index-scheduler/src/utils.rs

@@ -3,7 +3,7 @@
 use std::collections::{BTreeSet, HashSet};
 use std::ops::Bound;
 
-use meilisearch_types::batches::{Batch, BatchId, BatchStats};
+use meilisearch_types::batches::{Batch, BatchEnqueuedAt, BatchId, BatchStats};
 use meilisearch_types::heed::{Database, RoTxn, RwTxn};
 use meilisearch_types::milli::CboRoaringBitmapCodec;
 use meilisearch_types::task_view::DetailsView;
@@ -30,8 +30,7 @@ pub struct ProcessingBatch {
     pub kinds: HashSet<Kind>,
     pub indexes: HashSet<String>,
     pub canceled_by: HashSet<TaskId>,
-    pub oldest_enqueued_at: Option<OffsetDateTime>,
-    pub earliest_enqueued_at: Option<OffsetDateTime>,
+    pub enqueued_at: Option<BatchEnqueuedAt>,
     pub started_at: OffsetDateTime,
     pub finished_at: Option<OffsetDateTime>,
 }
@@ -51,8 +50,7 @@ impl ProcessingBatch {
             kinds: HashSet::default(),
             indexes: HashSet::default(),
             canceled_by: HashSet::default(),
-            oldest_enqueued_at: None,
-            earliest_enqueued_at: None,
+            enqueued_at: None,
             started_at: OffsetDateTime::now_utc(),
             finished_at: None,
         }
@@ -80,14 +78,18 @@ impl ProcessingBatch {
             if let Some(canceled_by) = task.canceled_by {
                 self.canceled_by.insert(canceled_by);
             }
-            self.oldest_enqueued_at =
-                Some(self.oldest_enqueued_at.map_or(task.enqueued_at, |oldest_enqueued_at| {
-                    task.enqueued_at.min(oldest_enqueued_at)
-                }));
-            self.earliest_enqueued_at =
-                Some(self.earliest_enqueued_at.map_or(task.enqueued_at, |earliest_enqueued_at| {
-                    task.enqueued_at.max(earliest_enqueued_at)
-                }));
+            match self.enqueued_at.as_mut() {
+                Some(BatchEnqueuedAt { earliest, oldest }) => {
+                    *oldest = task.enqueued_at.min(*oldest);
+                    *earliest = task.enqueued_at.max(*earliest);
+                }
+                None => {
+                    self.enqueued_at = Some(BatchEnqueuedAt {
+                        earliest: task.enqueued_at,
+                        oldest: task.enqueued_at,
+                    });
+                }
+            }
         }
     }
 
@@ -138,6 +140,7 @@ impl ProcessingBatch {
             stats: self.stats.clone(),
             started_at: self.started_at,
             finished_at: self.finished_at,
+            enqueued_at: self.enqueued_at,
         }
     }
 }
@@ -174,6 +177,33 @@ pub(crate) fn remove_task_datetime(
     Ok(())
 }
 
+pub(crate) fn remove_n_tasks_datetime_earlier_than(
+    wtxn: &mut RwTxn,
+    database: Database<BEI128, CboRoaringBitmapCodec>,
+    earlier_than: OffsetDateTime,
+    mut count: usize,
+    task_id: TaskId,
+) -> Result<()> {
+    let earlier_than = earlier_than.unix_timestamp_nanos();
+    let mut iter = database.rev_range_mut(wtxn, &(..earlier_than))?;
+    while let Some((current, mut existing)) = iter.next().transpose()? {
+        count -= existing.remove(task_id) as usize;
+
+        if existing.is_empty() {
+            // safety: We don't keep references to the database
+            unsafe { iter.del_current()? };
+        } else {
+            // safety: We don't keep references to the database
+            unsafe { iter.put_current(&current, &existing)? };
+        }
+        if count == 0 {
+            break;
+        }
+    }
+
+    Ok(())
+}
+
 pub(crate) fn keep_ids_within_datetimes(
     rtxn: &RoTxn,
     ids: &mut RoaringBitmap,
@@ -329,14 +359,27 @@ impl crate::IndexScheduler {
                 kind,
             } = task;
             assert_eq!(uid, task.uid);
-            if let Some(ref batch) = batch_uid {
+            if task.status != Status::Enqueued {
+                let batch_uid = batch_uid.expect("All non enqueued tasks must be part of a batch");
                 assert!(self
                     .queue
                     .batch_to_tasks_mapping
-                    .get(&rtxn, batch)
+                    .get(&rtxn, &batch_uid)
                     .unwrap()
                     .unwrap()
                     .contains(uid));
+                let batch = self.queue.batches.get_batch(&rtxn, batch_uid).unwrap().unwrap();
+                assert_eq!(batch.uid, batch_uid);
+                if task.status == Status::Processing {
+                    assert!(batch.progress.is_some());
+                } else {
+                    assert!(batch.progress.is_none());
+                }
+                assert_eq!(batch.started_at, task.started_at.unwrap());
+                assert_eq!(batch.finished_at, task.finished_at);
+                let enqueued_at = batch.enqueued_at.unwrap();
+                assert!(task.enqueued_at >= enqueued_at.oldest);
+                assert!(task.enqueued_at <= enqueued_at.earliest);
             }
             if let Some(task_index_uid) = &task_index_uid {
                 assert!(self

crates/index-scheduler/src/versioning.rs

@@ -1,9 +1,10 @@
-use crate::{upgrade::upgrade_index_scheduler, Result};
-use meilisearch_types::{
-    heed::{types::Str, Database, Env, RoTxn, RwTxn},
-    milli::heed_codec::version::VersionCodec,
-    versioning,
-};
+use meilisearch_types::heed::types::Str;
+use meilisearch_types::heed::{self, Database, Env, RoTxn, RwTxn};
+use meilisearch_types::milli::heed_codec::version::VersionCodec;
+use meilisearch_types::versioning;
+
+use crate::upgrade::upgrade_index_scheduler;
+use crate::Result;
 
 /// The number of database used by queue itself
 const NUMBER_OF_DATABASES: u32 = 1;
@@ -21,30 +22,38 @@ pub struct Versioning {
 }
 
 impl Versioning {
-    pub(crate) const fn nb_db() -> u32 {
+    pub const fn nb_db() -> u32 {
         NUMBER_OF_DATABASES
     }
 
-    pub fn get_version(&self, rtxn: &RoTxn) -> Result<Option<(u32, u32, u32)>> {
-        Ok(self.version.get(rtxn, entry_name::MAIN)?)
+    pub fn get_version(&self, rtxn: &RoTxn) -> Result<Option<(u32, u32, u32)>, heed::Error> {
+        self.version.get(rtxn, entry_name::MAIN)
     }
 
-    pub fn set_version(&self, wtxn: &mut RwTxn, version: (u32, u32, u32)) -> Result<()> {
-        Ok(self.version.put(wtxn, entry_name::MAIN, &version)?)
+    pub fn set_version(
+        &self,
+        wtxn: &mut RwTxn,
+        version: (u32, u32, u32),
+    ) -> Result<(), heed::Error> {
+        self.version.put(wtxn, entry_name::MAIN, &version)
     }
 
-    pub fn set_current_version(&self, wtxn: &mut RwTxn) -> Result<()> {
+    pub fn set_current_version(&self, wtxn: &mut RwTxn) -> Result<(), heed::Error> {
         let major = versioning::VERSION_MAJOR.parse().unwrap();
         let minor = versioning::VERSION_MINOR.parse().unwrap();
         let patch = versioning::VERSION_PATCH.parse().unwrap();
         self.set_version(wtxn, (major, minor, patch))
     }
 
-    /// Create an index scheduler and start its run loop.
+    /// Return `Self` without checking anything about the version
+    pub fn raw_new(env: &Env, wtxn: &mut RwTxn) -> Result<Self, heed::Error> {
+        let version = env.create_database(wtxn, Some(db_name::VERSION))?;
+        Ok(Self { version })
+    }
+
     pub(crate) fn new(env: &Env, db_version: (u32, u32, u32)) -> Result<Self> {
         let mut wtxn = env.write_txn()?;
-        let version = env.create_database(&mut wtxn, Some(db_name::VERSION))?;
-        let this = Self { version };
+        let this = Self::raw_new(env, &mut wtxn)?;
         let from = match this.get_version(&wtxn)? {
             Some(version) => version,
             // fresh DB: use the db version

crates/meilisearch-types/Cargo.toml

@@ -14,6 +14,7 @@ license.workspace = true
 actix-web = { version = "4.9.0", default-features = false }
 anyhow = "1.0.95"
 bumpalo = "3.16.0"
+bumparaw-collections = "0.1.4"
 convert_case = "0.6.0"
 csv = "1.3.1"
 deserr = { version = "0.6.3", features = ["actix-web"] }
@@ -24,12 +25,11 @@ flate2 = "1.0.35"
 fst = "0.4.7"
 memmap2 = "0.9.5"
 milli = { path = "../milli" }
-bumparaw-collections = "0.1.4"
 roaring = { version = "0.10.10", features = ["serde"] }
 rustc-hash = "2.1.0"
 serde = { version = "1.0.217", features = ["derive"] }
 serde-cs = "0.2.4"
-serde_json = "1.0.135"
+serde_json = { version = "1.0.135", features = ["preserve_order"] }
 tar = "0.4.43"
 tempfile = "3.15.0"
 thiserror = "2.0.9"

crates/meilisearch-types/src/batches.rs

@@ -24,9 +24,35 @@ pub struct Batch {
     pub started_at: OffsetDateTime,
     #[serde(with = "time::serde::rfc3339::option")]
     pub finished_at: Option<OffsetDateTime>,
+
+    // Enqueued at is never displayed and is only required when removing a batch.
+    // It's always some except when upgrading from a database pre v1.12
+    pub enqueued_at: Option<BatchEnqueuedAt>,
 }
 
-#[derive(Default, Debug, Clone, Serialize, Deserialize, ToSchema)]
+impl PartialEq for Batch {
+    fn eq(&self, other: &Self) -> bool {
+        let Self { uid, progress, details, stats, started_at, finished_at, enqueued_at } = self;
+
+        *uid == other.uid
+            && progress.is_none() == other.progress.is_none()
+            && details == &other.details
+            && stats == &other.stats
+            && started_at == &other.started_at
+            && finished_at == &other.finished_at
+            && enqueued_at == &other.enqueued_at
+    }
+}
+
+#[derive(Clone, Copy, Debug, PartialEq, Eq, Serialize, Deserialize)]
+pub struct BatchEnqueuedAt {
+    #[serde(with = "time::serde::rfc3339")]
+    pub earliest: OffsetDateTime,
+    #[serde(with = "time::serde::rfc3339")]
+    pub oldest: OffsetDateTime,
+}
+
+#[derive(Default, Debug, Clone, PartialEq, Eq, Serialize, Deserialize, ToSchema)]
 #[serde(rename_all = "camelCase")]
 #[schema(rename_all = "camelCase")]
 pub struct BatchStats {
@@ -34,4 +60,8 @@ pub struct BatchStats {
     pub status: BTreeMap<Status, u32>,
     pub types: BTreeMap<Kind, u32>,
     pub index_uids: BTreeMap<String, u32>,
+    #[serde(default, skip_serializing_if = "serde_json::Map::is_empty")]
+    pub progress_trace: serde_json::Map<String, serde_json::Value>,
+    #[serde(default, skip_serializing_if = "Option::is_none")]
+    pub write_channel_congestion: Option<serde_json::Map<String, serde_json::Value>>,
 }

crates/meilisearch-types/src/deserr/mod.rs

@@ -193,6 +193,8 @@ merge_with_error_impl_take_error_message!(ParseTaskKindError);
 merge_with_error_impl_take_error_message!(ParseTaskStatusError);
 merge_with_error_impl_take_error_message!(IndexUidFormatError);
 merge_with_error_impl_take_error_message!(InvalidMultiSearchWeight);
+merge_with_error_impl_take_error_message!(InvalidNetworkUrl);
+merge_with_error_impl_take_error_message!(InvalidNetworkSearchApiKey);
 merge_with_error_impl_take_error_message!(InvalidSearchSemanticRatio);
 merge_with_error_impl_take_error_message!(InvalidSearchRankingScoreThreshold);
 merge_with_error_impl_take_error_message!(InvalidSimilarRankingScoreThreshold);

crates/meilisearch-types/src/error.rs

@@ -260,7 +260,13 @@ InvalidMultiSearchMergeFacets         , InvalidRequest       , BAD_REQUEST ;
 InvalidMultiSearchQueryFacets         , InvalidRequest       , BAD_REQUEST ;
 InvalidMultiSearchQueryPagination     , InvalidRequest       , BAD_REQUEST ;
 InvalidMultiSearchQueryRankingRules   , InvalidRequest       , BAD_REQUEST ;
+InvalidMultiSearchQueryPosition       , InvalidRequest       , BAD_REQUEST ;
+InvalidMultiSearchRemote              , InvalidRequest       , BAD_REQUEST ;
 InvalidMultiSearchWeight              , InvalidRequest       , BAD_REQUEST ;
+InvalidNetworkRemotes                 , InvalidRequest       , BAD_REQUEST ;
+InvalidNetworkSelf                    , InvalidRequest       , BAD_REQUEST ;
+InvalidNetworkSearchApiKey            , InvalidRequest       , BAD_REQUEST ;
+InvalidNetworkUrl                     , InvalidRequest       , BAD_REQUEST ;
 InvalidSearchAttributesToSearchOn     , InvalidRequest       , BAD_REQUEST ;
 InvalidSearchAttributesToCrop         , InvalidRequest       , BAD_REQUEST ;
 InvalidSearchAttributesToHighlight    , InvalidRequest       , BAD_REQUEST ;
@@ -351,14 +357,22 @@ MissingDocumentId                     , InvalidRequest       , BAD_REQUEST ;
 MissingFacetSearchFacetName           , InvalidRequest       , BAD_REQUEST ;
 MissingIndexUid                       , InvalidRequest       , BAD_REQUEST ;
 MissingMasterKey                      , Auth                 , UNAUTHORIZED ;
+MissingNetworkUrl                     , InvalidRequest       , BAD_REQUEST ;
 MissingPayload                        , InvalidRequest       , BAD_REQUEST ;
 MissingSearchHybrid                   , InvalidRequest       , BAD_REQUEST ;
 MissingSwapIndexes                    , InvalidRequest       , BAD_REQUEST ;
 MissingTaskFilters                    , InvalidRequest       , BAD_REQUEST ;
 NoSpaceLeftOnDevice                   , System               , UNPROCESSABLE_ENTITY;
 PayloadTooLarge                       , InvalidRequest       , PAYLOAD_TOO_LARGE ;
+RemoteBadResponse                     , System               , BAD_GATEWAY ;
+RemoteBadRequest                      , InvalidRequest       , BAD_REQUEST ;
+RemoteCouldNotSendRequest             , System               , BAD_GATEWAY ;
+RemoteInvalidApiKey                   , Auth                 , FORBIDDEN ;
+RemoteRemoteError                     , System               , BAD_GATEWAY ;
+RemoteTimeout                         , System               , BAD_GATEWAY ;
 TooManySearchRequests                 , System               , SERVICE_UNAVAILABLE ;
 TaskNotFound                          , InvalidRequest       , NOT_FOUND ;
+TaskFileNotFound                      , InvalidRequest       , NOT_FOUND ;
 BatchNotFound                         , InvalidRequest       , NOT_FOUND ;
 TooManyOpenFiles                      , System               , UNPROCESSABLE_ENTITY ;
 TooManyVectors                        , InvalidRequest       , BAD_REQUEST ;
@@ -400,6 +414,7 @@ impl ErrorCode for milli::Error {
                     UserError::AttributeLimitReached => Code::MaxFieldsLimitExceeded,
                     UserError::InvalidFilter(_) => Code::InvalidSearchFilter,
                     UserError::InvalidFilterExpression(..) => Code::InvalidSearchFilter,
+                    UserError::FilterOperatorNotAllowed { .. } => Code::InvalidSearchFilter,
                     UserError::MissingDocumentId { .. } => Code::MissingDocumentId,
                     UserError::InvalidDocumentId { .. } | UserError::TooManyDocumentIds { .. } => {
                         Code::InvalidDocumentId
@@ -414,9 +429,10 @@ impl ErrorCode for milli::Error {
                     | UserError::InvalidUrl { .. }
                     | UserError::InvalidSettingsDocumentTemplateMaxBytes { .. }
                     | UserError::InvalidPrompt(_)
-                    | UserError::InvalidDisableBinaryQuantization { .. } => {
-                        Code::InvalidSettingsEmbedders
-                    }
+                    | UserError::InvalidDisableBinaryQuantization { .. }
+                    | UserError::InvalidSourceForNested { .. }
+                    | UserError::MissingSourceForNested { .. }
+                    | UserError::InvalidSettingsEmbedder { .. } => Code::InvalidSettingsEmbedders,
                     UserError::TooManyEmbedders(_) => Code::InvalidSettingsEmbedders,
                     UserError::InvalidPromptForEmbeddings(..) => Code::InvalidSettingsEmbedders,
                     UserError::NoPrimaryKeyCandidateFound => Code::IndexPrimaryKeyNoCandidateFound,
@@ -583,6 +599,18 @@ impl fmt::Display for deserr_codes::InvalidSimilarRankingScoreThreshold {
     }
 }
 
+impl fmt::Display for deserr_codes::InvalidNetworkUrl {
+    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+        write!(f, "the value of `url` is invalid, expected a string.")
+    }
+}
+
+impl fmt::Display for deserr_codes::InvalidNetworkSearchApiKey {
+    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+        write!(f, "the value of `searchApiKey` is invalid, expected a string.")
+    }
+}
+
 #[macro_export]
 macro_rules! internal_error {
     ($target:ty : $($other:path), *) => {

crates/meilisearch-types/src/features.rs

@@ -1,3 +1,5 @@
+use std::collections::BTreeMap;
+
 use serde::{Deserialize, Serialize};
 
 #[derive(Serialize, Deserialize, Debug, Clone, Copy, Default, PartialEq, Eq)]
@@ -7,6 +9,8 @@ pub struct RuntimeTogglableFeatures {
     pub logs_route: bool,
     pub edit_documents_by_function: bool,
     pub contains_filter: bool,
+    pub network: bool,
+    pub get_task_documents_route: bool,
 }
 
 #[derive(Default, Debug, Clone, Copy)]
@@ -15,3 +19,20 @@ pub struct InstanceTogglableFeatures {
     pub logs_route: bool,
     pub contains_filter: bool,
 }
+
+#[derive(Serialize, Deserialize, Debug, Clone, PartialEq, Eq)]
+#[serde(rename_all = "camelCase")]
+pub struct Remote {
+    pub url: String,
+    #[serde(default)]
+    pub search_api_key: Option<String>,
+}
+
+#[derive(Serialize, Deserialize, Debug, Clone, PartialEq, Eq, Default)]
+#[serde(rename_all = "camelCase")]
+pub struct Network {
+    #[serde(default, rename = "self")]
+    pub local: Option<String>,
+    #[serde(default)]
+    pub remotes: BTreeMap<String, Remote>,
+}

crates/meilisearch-types/src/index_uid.rs

@@ -4,13 +4,14 @@ use std::fmt;
 use std::str::FromStr;
 
 use deserr::Deserr;
+use serde::Serialize;
 use utoipa::ToSchema;
 
 use crate::error::{Code, ErrorCode};
 
 /// An index uid is composed of only ascii alphanumeric characters, - and _, between 1 and 400
 /// bytes long
-#[derive(Debug, Clone, PartialEq, Eq, Deserr, PartialOrd, Ord, ToSchema)]
+#[derive(Debug, Clone, PartialEq, Eq, Deserr, PartialOrd, Ord, Serialize, ToSchema)]
 #[deserr(try_from(String) = IndexUid::try_from -> IndexUidFormatError)]
 #[schema(value_type = String, example = "movies")]
 pub struct IndexUid(String);

crates/meilisearch-types/src/keys.rs

@@ -302,6 +302,12 @@ pub enum Action {
     #[serde(rename = "experimental.update")]
     #[deserr(rename = "experimental.update")]
     ExperimentalFeaturesUpdate,
+    #[serde(rename = "network.get")]
+    #[deserr(rename = "network.get")]
+    NetworkGet,
+    #[serde(rename = "network.update")]
+    #[deserr(rename = "network.update")]
+    NetworkUpdate,
 }
 
 impl Action {
@@ -341,6 +347,8 @@ impl Action {
             KEYS_DELETE => Some(Self::KeysDelete),
             EXPERIMENTAL_FEATURES_GET => Some(Self::ExperimentalFeaturesGet),
             EXPERIMENTAL_FEATURES_UPDATE => Some(Self::ExperimentalFeaturesUpdate),
+            NETWORK_GET => Some(Self::NetworkGet),
+            NETWORK_UPDATE => Some(Self::NetworkUpdate),
             _otherwise => None,
         }
     }
@@ -386,4 +394,7 @@ pub mod actions {
     pub const KEYS_DELETE: u8 = KeysDelete.repr();
     pub const EXPERIMENTAL_FEATURES_GET: u8 = ExperimentalFeaturesGet.repr();
     pub const EXPERIMENTAL_FEATURES_UPDATE: u8 = ExperimentalFeaturesUpdate.repr();
+
+    pub const NETWORK_GET: u8 = NetworkGet.repr();
+    pub const NETWORK_UPDATE: u8 = NetworkUpdate.repr();
 }

crates/meilisearch-types/src/locales.rs

@@ -1,5 +1,5 @@
 use deserr::Deserr;
-use milli::LocalizedAttributesRule;
+use milli::{AttributePatterns, LocalizedAttributesRule};
 use serde::{Deserialize, Serialize};
 use utoipa::ToSchema;
 
@@ -7,7 +7,7 @@ use utoipa::ToSchema;
 #[deserr(rename_all = camelCase)]
 #[serde(rename_all = "camelCase")]
 pub struct LocalizedAttributesRuleView {
-    pub attribute_patterns: Vec<String>,
+    pub attribute_patterns: AttributePatterns,
     pub locales: Vec<Locale>,
 }
 

crates/meilisearch-types/src/settings.rs

@@ -11,7 +11,7 @@ use fst::IntoStreamer;
 use milli::index::{IndexEmbeddingConfig, PrefixSearch};
 use milli::proximity::ProximityPrecision;
 use milli::update::Setting;
-use milli::{Criterion, CriterionError, Index, DEFAULT_VALUES_PER_FACET};
+use milli::{Criterion, CriterionError, FilterableAttributesRule, Index, DEFAULT_VALUES_PER_FACET};
 use serde::{Deserialize, Serialize, Serializer};
 use utoipa::ToSchema;
 
@@ -202,8 +202,8 @@ pub struct Settings<T> {
     /// Attributes to use for faceting and filtering. See [Filtering and Faceted Search](https://www.meilisearch.com/docs/learn/filtering_and_sorting/search_with_facet_filters).
     #[serde(default, skip_serializing_if = "Setting::is_not_set")]
     #[deserr(default, error = DeserrJsonError<InvalidSettingsFilterableAttributes>)]
-    #[schema(value_type = Option<Vec<String>>, example = json!(["release_date", "genre"]))]
-    pub filterable_attributes: Setting<BTreeSet<String>>,
+    #[schema(value_type = Option<Vec<FilterableAttributesRule>>, example = json!(["release_date", "genre"]))]
+    pub filterable_attributes: Setting<Vec<FilterableAttributesRule>>,
     /// Attributes to use when sorting search results.
     #[serde(default, skip_serializing_if = "Setting::is_not_set")]
     #[deserr(default, error = DeserrJsonError<InvalidSettingsSortableAttributes>)]
@@ -791,7 +791,7 @@ pub fn settings(
         .user_defined_searchable_fields(rtxn)?
         .map(|fields| fields.into_iter().map(String::from).collect());
 
-    let filterable_attributes = index.filterable_fields(rtxn)?.into_iter().collect();
+    let filterable_attributes = index.filterable_attributes_rules(rtxn)?.into_iter().collect();
 
     let sortable_attributes = index.sortable_fields(rtxn)?.into_iter().collect();
 

crates/meilisearch-types/src/versioning.rs

@@ -1,7 +1,10 @@
 use std::fs;
-use std::io::{self, ErrorKind};
+use std::io::{ErrorKind, Write};
 use std::path::Path;
 
+use milli::heed;
+use tempfile::NamedTempFile;
+
 /// The name of the file that contains the version of the database.
 pub const VERSION_FILE_NAME: &str = "VERSION";
 
@@ -10,37 +13,7 @@ pub static VERSION_MINOR: &str = env!("CARGO_PKG_VERSION_MINOR");
 pub static VERSION_PATCH: &str = env!("CARGO_PKG_VERSION_PATCH");
 
 /// Persists the version of the current Meilisearch binary to a VERSION file
-pub fn update_version_file_for_dumpless_upgrade(
-    db_path: &Path,
-    from: (u32, u32, u32),
-    to: (u32, u32, u32),
-) -> Result<(), VersionFileError> {
-    let (from_major, from_minor, from_patch) = from;
-    let (to_major, to_minor, to_patch) = to;
-
-    if from_major > to_major
-        || (from_major == to_major && from_minor > to_minor)
-        || (from_major == to_major && from_minor == to_minor && from_patch > to_patch)
-    {
-        Err(VersionFileError::DowngradeNotSupported {
-            major: from_major,
-            minor: from_minor,
-            patch: from_patch,
-        })
-    } else if from_major < 1 || (from_major == to_major && from_minor < 12) {
-        Err(VersionFileError::TooOldForAutomaticUpgrade {
-            major: from_major,
-            minor: from_minor,
-            patch: from_patch,
-        })
-    } else {
-        create_current_version_file(db_path)?;
-        Ok(())
-    }
-}
-
-/// Persists the version of the current Meilisearch binary to a VERSION file
-pub fn create_current_version_file(db_path: &Path) -> io::Result<()> {
+pub fn create_current_version_file(db_path: &Path) -> anyhow::Result<()> {
     create_version_file(db_path, VERSION_MAJOR, VERSION_MINOR, VERSION_PATCH)
 }
 
@@ -49,9 +22,14 @@ pub fn create_version_file(
     major: &str,
     minor: &str,
     patch: &str,
-) -> io::Result<()> {
+) -> anyhow::Result<()> {
     let version_path = db_path.join(VERSION_FILE_NAME);
-    fs::write(version_path, format!("{}.{}.{}", major, minor, patch))
+    // In order to persist the file later we must create it in the `data.ms` and not in `/tmp`
+    let mut file = NamedTempFile::new_in(db_path)?;
+    file.write_all(format!("{}.{}.{}", major, minor, patch).as_bytes())?;
+    file.flush()?;
+    file.persist(version_path)?;
+    Ok(())
 }
 
 pub fn get_version(db_path: &Path) -> Result<(u32, u32, u32), VersionFileError> {
@@ -61,7 +39,7 @@ pub fn get_version(db_path: &Path) -> Result<(u32, u32, u32), VersionFileError>
         Ok(version) => parse_version(&version),
         Err(error) => match error.kind() {
             ErrorKind::NotFound => Err(VersionFileError::MissingVersionFile),
-            _ => Err(error.into()),
+            _ => Err(anyhow::Error::from(error).into()),
         },
     }
 }
@@ -112,7 +90,9 @@ pub enum VersionFileError {
     DowngradeNotSupported { major: u32, minor: u32, patch: u32 },
     #[error("Database version {major}.{minor}.{patch} is too old for the experimental dumpless upgrade feature. Please generate a dump using the v{major}.{minor}.{patch} and import it in the v{VERSION_MAJOR}.{VERSION_MINOR}.{VERSION_PATCH}")]
     TooOldForAutomaticUpgrade { major: u32, minor: u32, patch: u32 },
+    #[error("Error while modifying the database: {0}")]
+    ErrorWhileModifyingTheDatabase(#[from] heed::Error),
 
     #[error(transparent)]
-    IoError(#[from] std::io::Error),
+    AnyhowError(#[from] anyhow::Error),
 }

crates/meilisearch/Cargo.toml

@@ -170,5 +170,5 @@ german = ["meilisearch-types/german"]
 turkish = ["meilisearch-types/turkish"]
 
 [package.metadata.mini-dashboard]
-assets-url = "https://github.com/meilisearch/mini-dashboard/releases/download/v0.2.16/build.zip"
-sha1 = "68f83438a114aabbe76bc9fe480071e741996662"
+assets-url = "https://github.com/meilisearch/mini-dashboard/releases/download/v0.2.18/build.zip"
+sha1 = "b408a30dcb6e20cddb0c153c23385bcac4c8e912"

crates/meilisearch/src/analytics/segment_analytics.rs

@@ -31,6 +31,7 @@ use crate::routes::{create_all_stats, Stats};
 use crate::Opt;
 
 const ANALYTICS_HEADER: &str = "X-Meilisearch-Client";
+const MEILI_SERVER_PROVIDER: &str = "MEILI_SERVER_PROVIDER";
 
 /// Write the instance-uid in the `data.ms` and in `~/.config/MeiliSearch/path-to-db-instance-uid`. Ignore the errors.
 fn write_user_id(db_path: &Path, user_id: &InstanceUid) {
@@ -195,6 +196,8 @@ struct Infos {
     experimental_reduce_indexing_memory_usage: bool,
     experimental_max_number_of_batched_tasks: usize,
     experimental_limit_batched_tasks_total_size: u64,
+    experimental_network: bool,
+    experimental_get_task_documents_route: bool,
     gpu_enabled: bool,
     db_path: bool,
     import_dump: bool,
@@ -285,6 +288,8 @@ impl Infos {
             logs_route,
             edit_documents_by_function,
             contains_filter,
+            network,
+            get_task_documents_route,
         } = features;
 
         // We're going to override every sensible information.
@@ -302,6 +307,8 @@ impl Infos {
             experimental_replication_parameters,
             experimental_enable_logs_route: experimental_enable_logs_route | logs_route,
             experimental_reduce_indexing_memory_usage,
+            experimental_network: network,
+            experimental_get_task_documents_route: get_task_documents_route,
             gpu_enabled: meilisearch_types::milli::vector::is_cuda_enabled(),
             db_path: db_path != PathBuf::from("./data.ms"),
             import_dump: import_dump.is_some(),
@@ -357,7 +364,7 @@ impl Segment {
                     "cores": sys.cpus().len(),
                     "ram_size": sys.total_memory(),
                     "disk_size": disks.iter().map(|disk| disk.total_space()).max(),
-                    "server_provider": std::env::var("MEILI_SERVER_PROVIDER").ok(),
+                    "server_provider": std::env::var(MEILI_SERVER_PROVIDER).ok(),
             })
         });
         let number_of_documents =
@@ -380,10 +387,18 @@ impl Segment {
         index_scheduler: Arc<IndexScheduler>,
         auth_controller: Arc<AuthController>,
     ) {
-        const INTERVAL: Duration = Duration::from_secs(60 * 60); // one hour
-                                                                 // The first batch must be sent after one hour.
+        let interval: Duration = match std::env::var(MEILI_SERVER_PROVIDER) {
+            Ok(provider) if provider.starts_with("meili_cloud:") => {
+                Duration::from_secs(60 * 60) // one hour
+            }
+            _ => {
+                // We're an open source instance
+                Duration::from_secs(60 * 60 * 24) // one day
+            }
+        };
+
         let mut interval =
-            tokio::time::interval_at(tokio::time::Instant::now() + INTERVAL, INTERVAL);
+            tokio::time::interval_at(tokio::time::Instant::now() + interval, interval);
 
         loop {
             select! {

crates/meilisearch/src/lib.rs

@@ -32,6 +32,7 @@ use analytics::Analytics;
 use anyhow::bail;
 use error::PayloadError;
 use extractors::payload::PayloadConfig;
+use index_scheduler::versioning::Versioning;
 use index_scheduler::{IndexScheduler, IndexSchedulerOptions};
 use meilisearch_auth::AuthController;
 use meilisearch_types::milli::constants::VERSION_MAJOR;
@@ -40,10 +41,9 @@ use meilisearch_types::milli::update::{IndexDocumentsConfig, IndexDocumentsMetho
 use meilisearch_types::settings::apply_settings_to_builder;
 use meilisearch_types::tasks::KindWithContent;
 use meilisearch_types::versioning::{
-    create_current_version_file, get_version, update_version_file_for_dumpless_upgrade,
-    VersionFileError, VERSION_MINOR, VERSION_PATCH,
+    create_current_version_file, get_version, VersionFileError, VERSION_MINOR, VERSION_PATCH,
 };
-use meilisearch_types::{compression, milli, VERSION_FILE_NAME};
+use meilisearch_types::{compression, heed, milli, VERSION_FILE_NAME};
 pub use option::Opt;
 use option::ScheduleSnapshot;
 use search_queue::SearchQueue;
@@ -356,14 +356,19 @@ fn open_or_create_database_unchecked(
 
 /// Ensures Meilisearch version is compatible with the database, returns an error in case of version mismatch.
 /// Returns the version that was contained in the version file
-fn check_version(opt: &Opt, binary_version: (u32, u32, u32)) -> anyhow::Result<(u32, u32, u32)> {
+fn check_version(
+    opt: &Opt,
+    index_scheduler_opt: &IndexSchedulerOptions,
+    binary_version: (u32, u32, u32),
+) -> anyhow::Result<(u32, u32, u32)> {
     let (bin_major, bin_minor, bin_patch) = binary_version;
     let (db_major, db_minor, db_patch) = get_version(&opt.db_path)?;
 
-    if db_major != bin_major || db_minor != bin_minor || db_patch > bin_patch {
+    if db_major != bin_major || db_minor != bin_minor || db_patch != bin_patch {
         if opt.experimental_dumpless_upgrade {
             update_version_file_for_dumpless_upgrade(
-                &opt.db_path,
+                opt,
+                index_scheduler_opt,
                 (db_major, db_minor, db_patch),
                 (bin_major, bin_minor, bin_patch),
             )?;
@@ -380,6 +385,57 @@ fn check_version(opt: &Opt, binary_version: (u32, u32, u32)) -> anyhow::Result<(
     Ok((db_major, db_minor, db_patch))
 }
 
+/// Persists the version of the current Meilisearch binary to a VERSION file
+pub fn update_version_file_for_dumpless_upgrade(
+    opt: &Opt,
+    index_scheduler_opt: &IndexSchedulerOptions,
+    from: (u32, u32, u32),
+    to: (u32, u32, u32),
+) -> Result<(), VersionFileError> {
+    let (from_major, from_minor, from_patch) = from;
+    let (to_major, to_minor, to_patch) = to;
+
+    // Early exit in case of error
+    if from_major > to_major
+        || (from_major == to_major && from_minor > to_minor)
+        || (from_major == to_major && from_minor == to_minor && from_patch > to_patch)
+    {
+        return Err(VersionFileError::DowngradeNotSupported {
+            major: from_major,
+            minor: from_minor,
+            patch: from_patch,
+        });
+    } else if from_major < 1 || (from_major == to_major && from_minor < 12) {
+        return Err(VersionFileError::TooOldForAutomaticUpgrade {
+            major: from_major,
+            minor: from_minor,
+            patch: from_patch,
+        });
+    }
+
+    // In the case of v1.12, the index-scheduler didn't store its internal version at the time.
+    // => We must write it immediately **in the index-scheduler** otherwise we'll update the version file
+    //    there is a risk of DB corruption if a restart happens after writing the version file but before
+    //    writing the version in the index-scheduler. See <https://github.com/meilisearch/meilisearch/issues/5280>
+    if from_major == 1 && from_minor == 12 {
+        let env = unsafe {
+            heed::EnvOpenOptions::new()
+                .max_dbs(Versioning::nb_db())
+                .map_size(index_scheduler_opt.task_db_size)
+                .open(&index_scheduler_opt.tasks_path)
+        }?;
+        let mut wtxn = env.write_txn()?;
+        let versioning = Versioning::raw_new(&env, &mut wtxn)?;
+        versioning.set_version(&mut wtxn, (from_major, from_minor, from_patch))?;
+        wtxn.commit()?;
+        // Should be instant since we're the only one using the env
+        env.prepare_for_closing().wait();
+    }
+
+    create_current_version_file(&opt.db_path)?;
+    Ok(())
+}
+
 /// Ensure you're in a valid state and open the IndexScheduler + AuthController for you.
 fn open_or_create_database(
     opt: &Opt,
@@ -387,7 +443,11 @@ fn open_or_create_database(
     empty_db: bool,
     binary_version: (u32, u32, u32),
 ) -> anyhow::Result<(IndexScheduler, AuthController)> {
-    let version = if !empty_db { check_version(opt, binary_version)? } else { binary_version };
+    let version = if !empty_db {
+        check_version(opt, &index_scheduler_opt, binary_version)?
+    } else {
+        binary_version
+    };
 
     open_or_create_database_unchecked(opt, index_scheduler_opt, OnFailure::KeepDb, version)
 }
@@ -431,10 +491,13 @@ fn import_dump(
         keys.push(key);
     }
 
-    // 3. Import the runtime features.
+    // 3. Import the runtime features and network
     let features = dump_reader.features()?.unwrap_or_default();
     index_scheduler.put_runtime_features(features)?;
 
+    let network = dump_reader.network()?.cloned().unwrap_or_default();
+    index_scheduler.put_network(network)?;
+
     let indexer_config = index_scheduler.indexer_config();
 
     // /!\ The tasks must be imported AFTER importing the indexes or else the scheduler might
@@ -508,9 +571,15 @@ fn import_dump(
         index_scheduler.refresh_index_stats(&uid)?;
     }
 
+    // 5. Import the queue
     let mut index_scheduler_dump = index_scheduler.register_dumped_task()?;
+    // 5.1. Import the batches
+    for ret in dump_reader.batches()? {
+        let batch = ret?;
+        index_scheduler_dump.register_dumped_batch(batch)?;
+    }
 
-    // 5. Import the tasks.
+    // 5.2. Import the tasks
     for ret in dump_reader.tasks()? {
         let (task, file) = ret?;
         index_scheduler_dump.register_dumped_task(task, file)?;

crates/meilisearch/src/routes/features.rs

@@ -50,6 +50,8 @@ pub fn configure(cfg: &mut web::ServiceConfig) {
             logs_route: Some(false),
             edit_documents_by_function: Some(false),
             contains_filter: Some(false),
+            network: Some(false),
+            get_task_documents_route: Some(false),
         })),
         (status = 401, description = "The authorization header is missing", body = ResponseError, content_type = "application/json", example = json!(
             {
@@ -88,6 +90,10 @@ pub struct RuntimeTogglableFeatures {
     pub edit_documents_by_function: Option<bool>,
     #[deserr(default)]
     pub contains_filter: Option<bool>,
+    #[deserr(default)]
+    pub network: Option<bool>,
+    #[deserr(default)]
+    pub get_task_documents_route: Option<bool>,
 }
 
 impl From<meilisearch_types::features::RuntimeTogglableFeatures> for RuntimeTogglableFeatures {
@@ -97,6 +103,8 @@ impl From<meilisearch_types::features::RuntimeTogglableFeatures> for RuntimeTogg
             logs_route,
             edit_documents_by_function,
             contains_filter,
+            network,
+            get_task_documents_route,
         } = value;
 
         Self {
@@ -104,6 +112,8 @@ impl From<meilisearch_types::features::RuntimeTogglableFeatures> for RuntimeTogg
             logs_route: Some(logs_route),
             edit_documents_by_function: Some(edit_documents_by_function),
             contains_filter: Some(contains_filter),
+            network: Some(network),
+            get_task_documents_route: Some(get_task_documents_route),
         }
     }
 }
@@ -114,6 +124,8 @@ pub struct PatchExperimentalFeatureAnalytics {
     logs_route: bool,
     edit_documents_by_function: bool,
     contains_filter: bool,
+    network: bool,
+    get_task_documents_route: bool,
 }
 
 impl Aggregate for PatchExperimentalFeatureAnalytics {
@@ -127,6 +139,8 @@ impl Aggregate for PatchExperimentalFeatureAnalytics {
             logs_route: new.logs_route,
             edit_documents_by_function: new.edit_documents_by_function,
             contains_filter: new.contains_filter,
+            network: new.network,
+            get_task_documents_route: new.get_task_documents_route,
         })
     }
 
@@ -149,6 +163,8 @@ impl Aggregate for PatchExperimentalFeatureAnalytics {
             logs_route: Some(false),
             edit_documents_by_function: Some(false),
             contains_filter: Some(false),
+            network: Some(false),
+            get_task_documents_route: Some(false),
          })),
         (status = 401, description = "The authorization header is missing", body = ResponseError, content_type = "application/json", example = json!(
             {
@@ -181,16 +197,23 @@ async fn patch_features(
             .edit_documents_by_function
             .unwrap_or(old_features.edit_documents_by_function),
         contains_filter: new_features.0.contains_filter.unwrap_or(old_features.contains_filter),
+        network: new_features.0.network.unwrap_or(old_features.network),
+        get_task_documents_route: new_features
+            .0
+            .get_task_documents_route
+            .unwrap_or(old_features.get_task_documents_route),
     };
 
     // explicitly destructure for analytics rather than using the `Serialize` implementation, because
-    // the it renames to camelCase, which we don't want for analytics.
+    // it renames to camelCase, which we don't want for analytics.
     // **Do not** ignore fields with `..` or `_` here, because we want to add them in the future.
     let meilisearch_types::features::RuntimeTogglableFeatures {
         metrics,
         logs_route,
         edit_documents_by_function,
         contains_filter,
+        network,
+        get_task_documents_route,
     } = new_features;
 
     analytics.publish(
@@ -199,6 +222,8 @@ async fn patch_features(
             logs_route,
             edit_documents_by_function,
             contains_filter,
+            network,
+            get_task_documents_route,
         },
         &req,
     );

crates/meilisearch/src/routes/indexes/mod.rs

@@ -494,8 +494,18 @@ pub async fn delete_index(
 pub struct IndexStats {
     /// Number of documents in the index
     pub number_of_documents: u64,
+    /// Size of the documents database, in bytes.
+    pub raw_document_db_size: u64,
+    /// Average size of a document in the documents database.
+    pub avg_document_size: u64,
     /// Whether or not the index is currently ingesting document
     pub is_indexing: bool,
+    /// Number of embeddings in the index
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub number_of_embeddings: Option<u64>,
+    /// Number of embedded documents in the index
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub number_of_embedded_documents: Option<u64>,
     /// Association of every field name with the number of times it occurs in the documents.
     #[schema(value_type = HashMap<String, u64>)]
     pub field_distribution: FieldDistribution,
@@ -504,8 +514,15 @@ pub struct IndexStats {
 impl From<index_scheduler::IndexStats> for IndexStats {
     fn from(stats: index_scheduler::IndexStats) -> Self {
         IndexStats {
-            number_of_documents: stats.inner_stats.number_of_documents,
+            number_of_documents: stats
+                .inner_stats
+                .number_of_documents
+                .unwrap_or(stats.inner_stats.documents_database_stats.number_of_entries()),
+            raw_document_db_size: stats.inner_stats.documents_database_stats.total_value_size(),
+            avg_document_size: stats.inner_stats.documents_database_stats.average_value_size(),
             is_indexing: stats.is_indexing,
+            number_of_embeddings: stats.inner_stats.number_of_embeddings,
+            number_of_embedded_documents: stats.inner_stats.number_of_embedded_documents,
             field_distribution: stats.inner_stats.field_distribution,
         }
     }
@@ -524,6 +541,10 @@ impl From<index_scheduler::IndexStats> for IndexStats {
         (status = OK, description = "The stats of the index", body = IndexStats, content_type = "application/json", example = json!(
             {
                 "numberOfDocuments": 10,
+                "rawDocumentDbSize": 10,
+                "avgDocumentSize": 10,
+                "numberOfEmbeddings": 10,
+                "numberOfEmbeddedDocuments": 10,
                 "isIndexing": true,
                 "fieldDistribution": {
                     "genre": 10,

crates/meilisearch/src/routes/indexes/settings.rs

@@ -291,7 +291,7 @@ make_setting_routes!(
     {
         route: "/filterable-attributes",
         update_verb: put,
-        value_type: std::collections::BTreeSet<String>,
+        value_type: Vec<meilisearch_types::milli::FilterableAttributesRule>,
         err_type: meilisearch_types::deserr::DeserrJsonError<
             meilisearch_types::error::deserr_codes::InvalidSettingsFilterableAttributes,
         >,

crates/meilisearch/src/routes/indexes/settings_analytics.rs

@@ -8,6 +8,7 @@ use std::collections::{BTreeMap, BTreeSet, HashSet};
 use meilisearch_types::facet_values_sort::FacetValuesSort;
 use meilisearch_types::locales::{Locale, LocalizedAttributesRuleView};
 use meilisearch_types::milli::update::Setting;
+use meilisearch_types::milli::FilterableAttributesRule;
 use meilisearch_types::settings::{
     FacetingSettings, PaginationSettings, PrefixSearchSettings, ProximityPrecisionView,
     RankingRuleView, SettingEmbeddingSettings, TypoSettings,
@@ -89,6 +90,10 @@ impl Aggregate for SettingsAnalytics {
             filterable_attributes: FilterableAttributesAnalytics {
                 total: new.filterable_attributes.total.or(self.filterable_attributes.total),
                 has_geo: new.filterable_attributes.has_geo.or(self.filterable_attributes.has_geo),
+                has_patterns: new
+                    .filterable_attributes
+                    .has_patterns
+                    .or(self.filterable_attributes.has_patterns),
             },
             distinct_attribute: DistinctAttributeAnalytics {
                 set: self.distinct_attribute.set | new.distinct_attribute.set,
@@ -328,13 +333,19 @@ impl SortableAttributesAnalytics {
 pub struct FilterableAttributesAnalytics {
     pub total: Option<usize>,
     pub has_geo: Option<bool>,
+    pub has_patterns: Option<bool>,
 }
 
 impl FilterableAttributesAnalytics {
-    pub fn new(setting: Option<&BTreeSet<String>>) -> Self {
+    pub fn new(setting: Option<&Vec<FilterableAttributesRule>>) -> Self {
         Self {
             total: setting.as_ref().map(|filter| filter.len()),
-            has_geo: setting.as_ref().map(|filter| filter.contains("_geo")),
+            has_geo: setting
+                .as_ref()
+                .map(|filter| filter.iter().any(FilterableAttributesRule::has_geo)),
+            has_patterns: setting.as_ref().map(|filter| {
+                filter.iter().any(|rule| matches!(rule, FilterableAttributesRule::Pattern(_)))
+            }),
         }
     }
 
@@ -512,6 +523,7 @@ impl EmbeddersAnalytics {
                     EmbedderSource::UserProvided => sources.insert("userProvided".to_string()),
                     EmbedderSource::Ollama => sources.insert("ollama".to_string()),
                     EmbedderSource::Rest => sources.insert("rest".to_string()),
+                    EmbedderSource::Composite => sources.insert("composite".to_string()),
                 };
             }
         };

crates/meilisearch/src/routes/mod.rs

@@ -9,6 +9,10 @@ use meilisearch_types::batches::BatchStats;
 use meilisearch_types::error::{Code, ErrorType, ResponseError};
 use meilisearch_types::index_uid::IndexUid;
 use meilisearch_types::keys::CreateApiKey;
+use meilisearch_types::milli::{
+    AttributePatterns, FilterFeatures, FilterableAttributesFeatures, FilterableAttributesPatterns,
+    FilterableAttributesRule,
+};
 use meilisearch_types::settings::{
     Checked, FacetingSettings, MinWordSizeTyposSetting, PaginationSettings, Settings, TypoSettings,
     Unchecked,
@@ -34,6 +38,7 @@ use crate::routes::features::RuntimeTogglableFeatures;
 use crate::routes::indexes::documents::{DocumentDeletionByFilter, DocumentEditionByFunction};
 use crate::routes::indexes::IndexView;
 use crate::routes::multi_search::SearchResults;
+use crate::routes::network::{Network, Remote};
 use crate::routes::swap_indexes::SwapIndexesPayload;
 use crate::search::{
     FederatedSearch, FederatedSearchResult, Federation, FederationOptions, MergeFacets,
@@ -54,6 +59,7 @@ mod logs;
 mod metrics;
 mod multi_search;
 mod multi_search_analytics;
+pub mod network;
 mod open_api_utils;
 mod snapshot;
 mod swap_indexes;
@@ -75,6 +81,7 @@ pub mod tasks;
         (path = "/multi-search", api = multi_search::MultiSearchApi),
         (path = "/swap-indexes", api = swap_indexes::SwapIndexesApi),
         (path = "/experimental-features", api = features::ExperimentalFeaturesApi),
+        (path = "/network", api = network::NetworkApi),
     ),
     paths(get_health, get_version, get_stats),
     tags(
@@ -85,7 +92,7 @@ pub mod tasks;
         url = "/",
         description = "Local server",
     )),
-    components(schemas(PaginationView<KeyView>, PaginationView<IndexView>, IndexView, DocumentDeletionByFilter, AllBatches, BatchStats, ProgressStepView, ProgressView, BatchView, RuntimeTogglableFeatures, SwapIndexesPayload, DocumentEditionByFunction, MergeFacets, FederationOptions, SearchQueryWithIndex, Federation, FederatedSearch, FederatedSearchResult, SearchResults, SearchResultWithIndex, SimilarQuery, SimilarResult, PaginationView<serde_json::Value>, BrowseQuery, UpdateIndexRequest, IndexUid, IndexCreateRequest, KeyView, Action, CreateApiKey, UpdateStderrLogs, LogMode, GetLogs, IndexStats, Stats, HealthStatus, HealthResponse, VersionResponse, Code, ErrorType, AllTasks, TaskView, Status, DetailsView, ResponseError, Settings<Unchecked>, Settings<Checked>, TypoSettings, MinWordSizeTyposSetting, FacetingSettings, PaginationSettings, SummarizedTaskView, Kind))
+    components(schemas(PaginationView<KeyView>, PaginationView<IndexView>, IndexView, DocumentDeletionByFilter, AllBatches, BatchStats, ProgressStepView, ProgressView, BatchView, RuntimeTogglableFeatures, SwapIndexesPayload, DocumentEditionByFunction, MergeFacets, FederationOptions, SearchQueryWithIndex, Federation, FederatedSearch, FederatedSearchResult, SearchResults, SearchResultWithIndex, SimilarQuery, SimilarResult, PaginationView<serde_json::Value>, BrowseQuery, UpdateIndexRequest, IndexUid, IndexCreateRequest, KeyView, Action, CreateApiKey, UpdateStderrLogs, LogMode, GetLogs, IndexStats, Stats, HealthStatus, HealthResponse, VersionResponse, Code, ErrorType, AllTasks, TaskView, Status, DetailsView, ResponseError, Settings<Unchecked>, Settings<Checked>, TypoSettings, MinWordSizeTyposSetting, FacetingSettings, PaginationSettings, SummarizedTaskView, Kind, Network, Remote, FilterableAttributesRule, FilterableAttributesPatterns, AttributePatterns, FilterableAttributesFeatures, FilterFeatures))
 )]
 pub struct MeilisearchApi;
 
@@ -103,7 +110,8 @@ pub fn configure(cfg: &mut web::ServiceConfig) {
         .service(web::scope("/multi-search").configure(multi_search::configure))
         .service(web::scope("/swap-indexes").configure(swap_indexes::configure))
         .service(web::scope("/metrics").configure(metrics::configure))
-        .service(web::scope("/experimental-features").configure(features::configure));
+        .service(web::scope("/experimental-features").configure(features::configure))
+        .service(web::scope("/network").configure(network::configure));
 
     #[cfg(feature = "swagger")]
     {
@@ -359,9 +367,9 @@ pub async fn running() -> HttpResponse {
 #[derive(Serialize, Debug, ToSchema)]
 #[serde(rename_all = "camelCase")]
 pub struct Stats {
-    /// The size of the database, in bytes.
+    /// The disk space used by the database, in bytes.
     pub database_size: u64,
-    #[serde(skip)]
+    /// The size of the database, in bytes.
     pub used_database_size: u64,
     /// The date of the last update in the RFC 3339 formats. Can be `null` if no update has ever been processed.
     #[serde(serialize_with = "time::serde::rfc3339::option::serialize")]
@@ -383,10 +391,14 @@ pub struct Stats {
         (status = 200, description = "The stats of the instance", body = Stats, content_type = "application/json", example = json!(
             {
                 "databaseSize": 567,
+                "usedDatabaseSize": 456,
                 "lastUpdate": "2019-11-20T09:40:33.711324Z",
                 "indexes": {
                     "movies": {
                         "numberOfDocuments": 10,
+                        "rawDocumentDbSize": 100,
+                        "maxDocumentSize": 16,
+                        "avgDocumentSize": 10,
                         "isIndexing": true,
                         "fieldDistribution": {
                             "genre": 10,

crates/meilisearch/src/routes/multi_search.rs

@@ -20,6 +20,7 @@ use crate::routes::indexes::search::search_kind;
 use crate::search::{
     add_search_rules, perform_federated_search, perform_search, FederatedSearch,
     FederatedSearchResult, RetrieveVectors, SearchQueryWithIndex, SearchResultWithIndex,
+    PROXY_SEARCH_HEADER, PROXY_SEARCH_HEADER_VALUE,
 };
 use crate::search_queue::SearchQueue;
 
@@ -48,6 +49,7 @@ pub struct SearchResults {
 /// Bundle multiple search queries in a single API request. Use this endpoint to search through multiple indexes at once.
 #[utoipa::path(
     post,
+    request_body = FederatedSearch,
     path = "",
     tag = "Multi-search",
     security(("Bearer" = ["search", "*"])),
@@ -186,18 +188,22 @@ pub async fn multi_search_with_post(
 
     let response = match federation {
         Some(federation) => {
-            let search_result = tokio::task::spawn_blocking(move || {
-                perform_federated_search(&index_scheduler, queries, federation, features)
-            })
-            .await;
+            // check remote header
+            let is_proxy = req
+                .headers()
+                .get(PROXY_SEARCH_HEADER)
+                .is_some_and(|value| value.as_bytes() == PROXY_SEARCH_HEADER_VALUE.as_bytes());
+            let search_result =
+                perform_federated_search(&index_scheduler, queries, federation, features, is_proxy)
+                    .await;
             permit.drop().await;
 
-            if let Ok(Ok(_)) = search_result {
+            if search_result.is_ok() {
                 multi_aggregate.succeed();
             }
 
             analytics.publish(multi_aggregate, &req);
-            HttpResponse::Ok().json(search_result??)
+            HttpResponse::Ok().json(search_result?)
         }
         None => {
             // Explicitly expect a `(ResponseError, usize)` for the error type rather than `ResponseError` only,

crates/meilisearch/src/routes/multi_search_analytics.rs

@@ -13,6 +13,8 @@ pub struct MultiSearchAggregator {
 
     // sum of the number of distinct indexes in each single request, use with total_received to compute an avg
     total_distinct_index_count: usize,
+    // sum of the number of distinct remotes in each single request, use with total_received to compute an avg
+    total_distinct_remote_count: usize,
     // number of queries with a single index, use with total_received to compute a proportion
     total_single_index: usize,
 
@@ -31,46 +33,49 @@ impl MultiSearchAggregator {
     pub fn from_federated_search(federated_search: &FederatedSearch) -> Self {
         let use_federation = federated_search.federation.is_some();
 
-        let distinct_indexes: HashSet<_> = federated_search
-            .queries
-            .iter()
-            .map(|query| {
-                let query = &query;
-                // make sure we get a compilation error if a field gets added to / removed from SearchQueryWithIndex
-                let SearchQueryWithIndex {
-                    index_uid,
-                    federation_options: _,
-                    q: _,
-                    vector: _,
-                    offset: _,
-                    limit: _,
-                    page: _,
-                    hits_per_page: _,
-                    attributes_to_retrieve: _,
-                    retrieve_vectors: _,
-                    attributes_to_crop: _,
-                    crop_length: _,
-                    attributes_to_highlight: _,
-                    show_ranking_score: _,
-                    show_ranking_score_details: _,
-                    show_matches_position: _,
-                    filter: _,
-                    sort: _,
-                    distinct: _,
-                    facets: _,
-                    highlight_pre_tag: _,
-                    highlight_post_tag: _,
-                    crop_marker: _,
-                    matching_strategy: _,
-                    attributes_to_search_on: _,
-                    hybrid: _,
-                    ranking_score_threshold: _,
-                    locales: _,
-                } = query;
+        let mut distinct_indexes = HashSet::with_capacity(federated_search.queries.len());
+        let mut distinct_remotes = HashSet::with_capacity(federated_search.queries.len());
 
-                index_uid.as_str()
-            })
-            .collect();
+        // make sure we get a compilation error if a field gets added to / removed from SearchQueryWithIndex
+        for SearchQueryWithIndex {
+            index_uid,
+            federation_options,
+            q: _,
+            vector: _,
+            offset: _,
+            limit: _,
+            page: _,
+            hits_per_page: _,
+            attributes_to_retrieve: _,
+            retrieve_vectors: _,
+            attributes_to_crop: _,
+            crop_length: _,
+            attributes_to_highlight: _,
+            show_ranking_score: _,
+            show_ranking_score_details: _,
+            show_matches_position: _,
+            filter: _,
+            sort: _,
+            distinct: _,
+            facets: _,
+            highlight_pre_tag: _,
+            highlight_post_tag: _,
+            crop_marker: _,
+            matching_strategy: _,
+            attributes_to_search_on: _,
+            hybrid: _,
+            ranking_score_threshold: _,
+            locales: _,
+        } in &federated_search.queries
+        {
+            if let Some(federation_options) = federation_options {
+                if let Some(remote) = &federation_options.remote {
+                    distinct_remotes.insert(remote.as_str());
+                }
+            }
+
+            distinct_indexes.insert(index_uid.as_str());
+        }
 
         let show_ranking_score =
             federated_search.queries.iter().any(|query| query.show_ranking_score);
@@ -81,6 +86,7 @@ impl MultiSearchAggregator {
             total_received: 1,
             total_succeeded: 0,
             total_distinct_index_count: distinct_indexes.len(),
+            total_distinct_remote_count: distinct_remotes.len(),
             total_single_index: if distinct_indexes.len() == 1 { 1 } else { 0 },
             total_search_count: federated_search.queries.len(),
             show_ranking_score,
@@ -110,6 +116,8 @@ impl Aggregate for MultiSearchAggregator {
         let total_succeeded = this.total_succeeded.saturating_add(new.total_succeeded);
         let total_distinct_index_count =
             this.total_distinct_index_count.saturating_add(new.total_distinct_index_count);
+        let total_distinct_remote_count =
+            this.total_distinct_remote_count.saturating_add(new.total_distinct_remote_count);
         let total_single_index = this.total_single_index.saturating_add(new.total_single_index);
         let total_search_count = this.total_search_count.saturating_add(new.total_search_count);
         let show_ranking_score = this.show_ranking_score || new.show_ranking_score;
@@ -121,6 +129,7 @@ impl Aggregate for MultiSearchAggregator {
             total_received,
             total_succeeded,
             total_distinct_index_count,
+            total_distinct_remote_count,
             total_single_index,
             total_search_count,
             show_ranking_score,
@@ -134,6 +143,7 @@ impl Aggregate for MultiSearchAggregator {
             total_received,
             total_succeeded,
             total_distinct_index_count,
+            total_distinct_remote_count,
             total_single_index,
             total_search_count,
             show_ranking_score,
@@ -152,6 +162,10 @@ impl Aggregate for MultiSearchAggregator {
                 "total_distinct_index_count": total_distinct_index_count,
                 "avg_distinct_index_count": (total_distinct_index_count as f64) / (total_received as f64), // not 0 else returned early
             },
+            "remotes": {
+                "total_distinct_remote_count": total_distinct_remote_count,
+                "avg_distinct_remote_count": (total_distinct_remote_count as f64) / (total_received as f64), // not 0 else returned early
+            },
             "searches": {
                 "total_search_count": total_search_count,
                 "avg_search_count": (total_search_count as f64) / (total_received as f64),

crates/meilisearch/src/routes/network.rs

@@ -0,0 +1,261 @@
+use std::collections::BTreeMap;
+
+use actix_web::web::{self, Data};
+use actix_web::{HttpRequest, HttpResponse};
+use deserr::actix_web::AwebJson;
+use deserr::Deserr;
+use index_scheduler::IndexScheduler;
+use itertools::{EitherOrBoth, Itertools};
+use meilisearch_types::deserr::DeserrJsonError;
+use meilisearch_types::error::deserr_codes::{
+    InvalidNetworkRemotes, InvalidNetworkSearchApiKey, InvalidNetworkSelf, InvalidNetworkUrl,
+};
+use meilisearch_types::error::ResponseError;
+use meilisearch_types::features::{Network as DbNetwork, Remote as DbRemote};
+use meilisearch_types::keys::actions;
+use meilisearch_types::milli::update::Setting;
+use serde::Serialize;
+use tracing::debug;
+use utoipa::{OpenApi, ToSchema};
+
+use crate::analytics::{Aggregate, Analytics};
+use crate::extractors::authentication::policies::ActionPolicy;
+use crate::extractors::authentication::GuardedData;
+use crate::extractors::sequential_extractor::SeqHandler;
+
+#[derive(OpenApi)]
+#[openapi(
+    paths(get_network, patch_network),
+    tags((
+        name = "Network",
+        description = "The `/network` route allows you to describe the topology of a network of Meilisearch instances.
+
+This route is **synchronous**. This means that no task object will be returned, and any change to the network will be made available immediately.",
+        external_docs(url = "https://www.meilisearch.com/docs/reference/api/network"),
+    )),
+)]
+pub struct NetworkApi;
+
+pub fn configure(cfg: &mut web::ServiceConfig) {
+    cfg.service(
+        web::resource("")
+            .route(web::get().to(get_network))
+            .route(web::patch().to(SeqHandler(patch_network))),
+    );
+}
+
+/// Get network topology
+///
+/// Get a list of all Meilisearch instances currently known to this instance.
+#[utoipa::path(
+    get,
+    path = "",
+    tag = "Network",
+    security(("Bearer" = ["network.get", "network.*", "*"])),
+    responses(
+        (status = OK, description = "Known nodes are returned", body = Network, content_type = "application/json", example = json!(
+            {
+            "self": "ms-0",
+            "remotes": {
+            "ms-0": Remote { url: Setting::Set("http://localhost:7700".into()), search_api_key: Setting::Reset },
+            "ms-1": Remote { url: Setting::Set("http://localhost:7701".into()), search_api_key: Setting::Set("foo".into()) },
+            "ms-2": Remote { url: Setting::Set("http://localhost:7702".into()), search_api_key: Setting::Set("bar".into()) },
+        }
+    })),
+        (status = 401, description = "The authorization header is missing", body = ResponseError, content_type = "application/json", example = json!(
+            {
+                "message": "The Authorization header is missing. It must use the bearer authorization method.",
+                "code": "missing_authorization_header",
+                "type": "auth",
+                "link": "https://docs.meilisearch.com/errors#missing_authorization_header"
+            }
+        )),
+    )
+)]
+async fn get_network(
+    index_scheduler: GuardedData<ActionPolicy<{ actions::NETWORK_GET }>, Data<IndexScheduler>>,
+) -> Result<HttpResponse, ResponseError> {
+    index_scheduler.features().check_network("Using the /network route")?;
+
+    let network = index_scheduler.network();
+    debug!(returns = ?network, "Get network");
+    Ok(HttpResponse::Ok().json(network))
+}
+
+#[derive(Debug, Deserr, ToSchema, Serialize)]
+#[deserr(error = DeserrJsonError<InvalidNetworkRemotes>, rename_all = camelCase, deny_unknown_fields)]
+#[serde(rename_all = "camelCase")]
+#[schema(rename_all = "camelCase")]
+pub struct Remote {
+    #[schema(value_type = Option<String>, example = json!({
+        "ms-0": Remote { url: Setting::Set("http://localhost:7700".into()), search_api_key: Setting::Reset },
+        "ms-1": Remote { url: Setting::Set("http://localhost:7701".into()), search_api_key: Setting::Set("foo".into()) },
+        "ms-2": Remote { url: Setting::Set("http://localhost:7702".into()), search_api_key: Setting::Set("bar".into()) },
+    }))]
+    #[deserr(default, error = DeserrJsonError<InvalidNetworkUrl>)]
+    #[serde(default)]
+    pub url: Setting<String>,
+    #[schema(value_type = Option<String>, example = json!("XWnBI8QHUc-4IlqbKPLUDuhftNq19mQtjc6JvmivzJU"))]
+    #[deserr(default, error = DeserrJsonError<InvalidNetworkSearchApiKey>)]
+    #[serde(default)]
+    pub search_api_key: Setting<String>,
+}
+
+#[derive(Debug, Deserr, ToSchema, Serialize)]
+#[deserr(error = DeserrJsonError, rename_all = camelCase, deny_unknown_fields)]
+#[serde(rename_all = "camelCase")]
+#[schema(rename_all = "camelCase")]
+pub struct Network {
+    #[schema(value_type = Option<BTreeMap<String, Remote>>, example = json!("http://localhost:7700"))]
+    #[deserr(default, error = DeserrJsonError<InvalidNetworkRemotes>)]
+    #[serde(default)]
+    pub remotes: Setting<BTreeMap<String, Option<Remote>>>,
+    #[schema(value_type = Option<String>, example = json!("ms-00"), rename = "self")]
+    #[serde(default, rename = "self")]
+    #[deserr(default, rename = "self", error = DeserrJsonError<InvalidNetworkSelf>)]
+    pub local: Setting<String>,
+}
+
+impl Remote {
+    pub fn try_into_db_node(self, name: &str) -> Result<DbRemote, ResponseError> {
+        Ok(DbRemote {
+            url: self.url.set().ok_or(ResponseError::from_msg(
+                format!("Missing field `.remotes.{name}.url`"),
+                meilisearch_types::error::Code::MissingNetworkUrl,
+            ))?,
+            search_api_key: self.search_api_key.set(),
+        })
+    }
+}
+
+#[derive(Serialize)]
+pub struct PatchNetworkAnalytics {
+    network_size: usize,
+    network_has_self: bool,
+}
+
+impl Aggregate for PatchNetworkAnalytics {
+    fn event_name(&self) -> &'static str {
+        "Network Updated"
+    }
+
+    fn aggregate(self: Box<Self>, new: Box<Self>) -> Box<Self> {
+        Box::new(Self { network_size: new.network_size, network_has_self: new.network_has_self })
+    }
+
+    fn into_event(self: Box<Self>) -> serde_json::Value {
+        serde_json::to_value(*self).unwrap_or_default()
+    }
+}
+
+/// Configure Network
+///
+/// Add or remove nodes from network.
+#[utoipa::path(
+    patch,
+    path = "",
+    tag = "Network",
+    request_body = Network,
+    security(("Bearer" = ["network.update", "network.*", "*"])),
+    responses(
+        (status = OK, description = "New network state is returned",  body = Network, content_type = "application/json", example = json!(
+            {
+                "self": "ms-0",
+                "remotes": {
+                "ms-0": Remote { url: Setting::Set("http://localhost:7700".into()), search_api_key: Setting::Reset },
+                "ms-1": Remote { url: Setting::Set("http://localhost:7701".into()), search_api_key: Setting::Set("foo".into()) },
+                "ms-2": Remote { url: Setting::Set("http://localhost:7702".into()), search_api_key: Setting::Set("bar".into()) },
+            }
+        })),
+        (status = 401, description = "The authorization header is missing", body = ResponseError, content_type = "application/json", example = json!(
+            {
+                "message": "The Authorization header is missing. It must use the bearer authorization method.",
+                "code": "missing_authorization_header",
+                "type": "auth",
+                "link": "https://docs.meilisearch.com/errors#missing_authorization_header"
+            }
+        )),
+    )
+)]
+async fn patch_network(
+    index_scheduler: GuardedData<ActionPolicy<{ actions::NETWORK_UPDATE }>, Data<IndexScheduler>>,
+    new_network: AwebJson<Network, DeserrJsonError>,
+    req: HttpRequest,
+    analytics: Data<Analytics>,
+) -> Result<HttpResponse, ResponseError> {
+    index_scheduler.features().check_network("Using the /network route")?;
+
+    let new_network = new_network.0;
+    let old_network = index_scheduler.network();
+    debug!(parameters = ?new_network, "Patch network");
+
+    let merged_self = match new_network.local {
+        Setting::Set(new_self) => Some(new_self),
+        Setting::Reset => None,
+        Setting::NotSet => old_network.local,
+    };
+
+    let merged_remotes = match new_network.remotes {
+        Setting::Set(new_remotes) => {
+            let mut merged_remotes = BTreeMap::new();
+            for either_or_both in old_network
+                .remotes
+                .into_iter()
+                .merge_join_by(new_remotes.into_iter(), |left, right| left.0.cmp(&right.0))
+            {
+                match either_or_both {
+                    EitherOrBoth::Both((key, old), (_, Some(new))) => {
+                        let DbRemote { url: old_url, search_api_key: old_search_api_key } = old;
+
+                        let Remote { url: new_url, search_api_key: new_search_api_key } = new;
+
+                        let merged = DbRemote {
+                            url: match new_url {
+                                Setting::Set(new_url) => new_url,
+                                Setting::Reset => {
+                                    return Err(ResponseError::from_msg(
+                                        format!(
+                                            "Field `.remotes.{key}.url` cannot be set to `null`"
+                                        ),
+                                        meilisearch_types::error::Code::InvalidNetworkUrl,
+                                    ))
+                                }
+                                Setting::NotSet => old_url,
+                            },
+                            search_api_key: match new_search_api_key {
+                                Setting::Set(new_search_api_key) => Some(new_search_api_key),
+                                Setting::Reset => None,
+                                Setting::NotSet => old_search_api_key,
+                            },
+                        };
+                        merged_remotes.insert(key, merged);
+                    }
+                    EitherOrBoth::Both((_, _), (_, None)) | EitherOrBoth::Right((_, None)) => {}
+                    EitherOrBoth::Left((key, node)) => {
+                        merged_remotes.insert(key, node);
+                    }
+                    EitherOrBoth::Right((key, Some(node))) => {
+                        let node = node.try_into_db_node(&key)?;
+                        merged_remotes.insert(key, node);
+                    }
+                }
+            }
+            merged_remotes
+        }
+        Setting::Reset => BTreeMap::new(),
+        Setting::NotSet => old_network.remotes,
+    };
+
+    analytics.publish(
+        PatchNetworkAnalytics {
+            network_size: merged_remotes.len(),
+            network_has_self: merged_self.is_some(),
+        },
+        &req,
+    );
+
+    let merged_network = DbNetwork { local: merged_self, remotes: merged_remotes };
+    index_scheduler.put_network(merged_network.clone())?;
+    debug!(returns = ?merged_network, "Patch network");
+    Ok(HttpResponse::Ok().json(merged_network))
+}

crates/meilisearch/src/routes/tasks.rs

@@ -1,3 +1,5 @@
+use std::io::ErrorKind;
+
 use actix_web::web::Data;
 use actix_web::{web, HttpRequest, HttpResponse};
 use deserr::actix_web::AwebQueryParameter;
@@ -16,6 +18,7 @@ use serde::Serialize;
 use time::format_description::well_known::Rfc3339;
 use time::macros::format_description;
 use time::{Date, Duration, OffsetDateTime, Time};
+use tokio::io::AsyncReadExt;
 use tokio::task;
 use utoipa::{IntoParams, OpenApi, ToSchema};
 
@@ -44,7 +47,11 @@ pub fn configure(cfg: &mut web::ServiceConfig) {
             .route(web::delete().to(SeqHandler(delete_tasks))),
     )
     .service(web::resource("/cancel").route(web::post().to(SeqHandler(cancel_tasks))))
-    .service(web::resource("/{task_id}").route(web::get().to(SeqHandler(get_task))));
+    .service(web::resource("/{task_id}").route(web::get().to(SeqHandler(get_task))))
+    .service(
+        web::resource("/{task_id}/documents")
+            .route(web::get().to(SeqHandler(get_task_documents_file))),
+    );
 }
 
 #[derive(Debug, Deserr, IntoParams)]
@@ -639,6 +646,76 @@ async fn get_task(
     }
 }
 
+/// Get a task's documents.
+///
+/// Get a [task's documents file](https://www.meilisearch.com/docs/learn/async/asynchronous_operations).
+#[utoipa::path(
+    get,
+    path = "/{taskUid}/documents",
+    tag = "Tasks",
+    security(("Bearer" = ["tasks.get", "tasks.*", "*"])),
+    params(("taskUid", format = UInt32, example = 0, description = "The task identifier", nullable = false)),
+    responses(
+        (status = 200, description = "The content of the task update", body = serde_json::Value, content_type = "application/x-ndjson"),
+        (status = 401, description = "The authorization header is missing", body = ResponseError, content_type = "application/json", example = json!(
+            {
+                "message": "The Authorization header is missing. It must use the bearer authorization method.",
+                "code": "missing_authorization_header",
+                "type": "auth",
+                "link": "https://docs.meilisearch.com/errors#missing_authorization_header"
+            }
+        )),
+        (status = 404, description = "The task uid does not exists", body = ResponseError, content_type = "application/json", example = json!(
+            {
+                "message": "Task :taskUid not found.",
+                "code": "task_not_found",
+                "type": "invalid_request",
+                "link": "https://docs.meilisearch.com/errors/#task_not_found"
+            }
+        ))
+    )
+)]
+async fn get_task_documents_file(
+    index_scheduler: GuardedData<ActionPolicy<{ actions::TASKS_GET }>, Data<IndexScheduler>>,
+    task_uid: web::Path<String>,
+) -> Result<HttpResponse, ResponseError> {
+    index_scheduler.features().check_get_task_documents_route()?;
+    let task_uid_string = task_uid.into_inner();
+
+    let task_uid: TaskId = match task_uid_string.parse() {
+        Ok(id) => id,
+        Err(_e) => {
+            return Err(index_scheduler::Error::InvalidTaskUid { task_uid: task_uid_string }.into())
+        }
+    };
+
+    let query = index_scheduler::Query { uids: Some(vec![task_uid]), ..Query::default() };
+    let filters = index_scheduler.filters();
+    let (tasks, _) = index_scheduler.get_tasks_from_authorized_indexes(&query, filters)?;
+
+    if let Some(task) = tasks.first() {
+        match task.content_uuid() {
+            Some(uuid) => {
+                let mut tfile = match index_scheduler.queue.update_file(uuid) {
+                    Ok(file) => tokio::fs::File::from_std(file),
+                    Err(file_store::Error::IoError(e)) if e.kind() == ErrorKind::NotFound => {
+                        return Err(index_scheduler::Error::TaskFileNotFound(task_uid).into())
+                    }
+                    Err(e) => return Err(e.into()),
+                };
+                // Yes, that's awful to put everything in memory when we could have streamed it from
+                // disk but it's really (really) complex to do with the current state of async Rust.
+                let mut content = String::new();
+                tfile.read_to_string(&mut content).await?;
+                Ok(HttpResponse::Ok().content_type("application/x-ndjson").body(content))
+            }
+            None => Err(index_scheduler::Error::TaskFileNotFound(task_uid).into()),
+        }
+    } else {
+        Err(index_scheduler::Error::TaskNotFound(task_uid).into())
+    }
+}
+
 pub enum DeserializeDateOption {
     Before,
     After,

crates/meilisearch/src/search/federated.rs

@@ -1,923 +0,0 @@
-use std::cmp::Ordering;
-use std::collections::BTreeMap;
-use std::fmt;
-use std::iter::Zip;
-use std::rc::Rc;
-use std::str::FromStr as _;
-use std::time::Duration;
-use std::vec::{IntoIter, Vec};
-
-use actix_http::StatusCode;
-use index_scheduler::{IndexScheduler, RoFeatures};
-use indexmap::IndexMap;
-use meilisearch_types::deserr::DeserrJsonError;
-use meilisearch_types::error::deserr_codes::{
-    InvalidMultiSearchFacetsByIndex, InvalidMultiSearchMaxValuesPerFacet,
-    InvalidMultiSearchMergeFacets, InvalidMultiSearchWeight, InvalidSearchLimit,
-    InvalidSearchOffset,
-};
-use meilisearch_types::error::ResponseError;
-use meilisearch_types::index_uid::IndexUid;
-use meilisearch_types::milli::score_details::{ScoreDetails, ScoreValue};
-use meilisearch_types::milli::{self, DocumentId, OrderBy, TimeBudget};
-use roaring::RoaringBitmap;
-use serde::Serialize;
-use utoipa::ToSchema;
-
-use super::ranking_rules::{self, RankingRules};
-use super::{
-    compute_facet_distribution_stats, prepare_search, AttributesFormat, ComputedFacets, FacetStats,
-    HitMaker, HitsInfo, RetrieveVectors, SearchHit, SearchKind, SearchQuery, SearchQueryWithIndex,
-};
-use crate::error::MeilisearchHttpError;
-use crate::routes::indexes::search::search_kind;
-
-pub const DEFAULT_FEDERATED_WEIGHT: f64 = 1.0;
-
-#[derive(Debug, Default, Clone, Copy, PartialEq, deserr::Deserr, ToSchema)]
-#[deserr(error = DeserrJsonError, rename_all = camelCase, deny_unknown_fields)]
-pub struct FederationOptions {
-    #[deserr(default, error = DeserrJsonError<InvalidMultiSearchWeight>)]
-    #[schema(value_type = f64)]
-    pub weight: Weight,
-}
-
-#[derive(Debug, Clone, Copy, PartialEq, deserr::Deserr)]
-#[deserr(try_from(f64) = TryFrom::try_from -> InvalidMultiSearchWeight)]
-pub struct Weight(f64);
-
-impl Default for Weight {
-    fn default() -> Self {
-        Weight(DEFAULT_FEDERATED_WEIGHT)
-    }
-}
-
-impl std::convert::TryFrom<f64> for Weight {
-    type Error = InvalidMultiSearchWeight;
-
-    fn try_from(f: f64) -> Result<Self, Self::Error> {
-        if f < 0.0 {
-            Err(InvalidMultiSearchWeight)
-        } else {
-            Ok(Weight(f))
-        }
-    }
-}
-
-impl std::ops::Deref for Weight {
-    type Target = f64;
-
-    fn deref(&self) -> &Self::Target {
-        &self.0
-    }
-}
-
-#[derive(Debug, deserr::Deserr, ToSchema)]
-#[deserr(error = DeserrJsonError, rename_all = camelCase, deny_unknown_fields)]
-#[schema(rename_all = "camelCase")]
-pub struct Federation {
-    #[deserr(default = super::DEFAULT_SEARCH_LIMIT(), error = DeserrJsonError<InvalidSearchLimit>)]
-    pub limit: usize,
-    #[deserr(default = super::DEFAULT_SEARCH_OFFSET(), error = DeserrJsonError<InvalidSearchOffset>)]
-    pub offset: usize,
-    #[deserr(default, error = DeserrJsonError<InvalidMultiSearchFacetsByIndex>)]
-    pub facets_by_index: BTreeMap<IndexUid, Option<Vec<String>>>,
-    #[deserr(default, error = DeserrJsonError<InvalidMultiSearchMergeFacets>)]
-    pub merge_facets: Option<MergeFacets>,
-}
-
-#[derive(Copy, Clone, Debug, deserr::Deserr, Default, ToSchema)]
-#[deserr(error = DeserrJsonError<InvalidMultiSearchMergeFacets>, rename_all = camelCase, deny_unknown_fields)]
-#[schema(rename_all = "camelCase")]
-pub struct MergeFacets {
-    #[deserr(default, error = DeserrJsonError<InvalidMultiSearchMaxValuesPerFacet>)]
-    pub max_values_per_facet: Option<usize>,
-}
-
-#[derive(Debug, deserr::Deserr, ToSchema)]
-#[deserr(error = DeserrJsonError, rename_all = camelCase, deny_unknown_fields)]
-#[schema(rename_all = "camelCase")]
-pub struct FederatedSearch {
-    pub queries: Vec<SearchQueryWithIndex>,
-    #[deserr(default)]
-    pub federation: Option<Federation>,
-}
-
-#[derive(Serialize, Clone, ToSchema)]
-#[serde(rename_all = "camelCase")]
-#[schema(rename_all = "camelCase")]
-pub struct FederatedSearchResult {
-    pub hits: Vec<SearchHit>,
-    pub processing_time_ms: u128,
-    #[serde(flatten)]
-    pub hits_info: HitsInfo,
-
-    #[serde(skip_serializing_if = "Option::is_none")]
-    pub semantic_hit_count: Option<u32>,
-
-    #[serde(skip_serializing_if = "Option::is_none")]
-    #[schema(value_type = Option<BTreeMap<String, BTreeMap<String, u64>>>)]
-    pub facet_distribution: Option<BTreeMap<String, IndexMap<String, u64>>>,
-    #[serde(skip_serializing_if = "Option::is_none")]
-    pub facet_stats: Option<BTreeMap<String, FacetStats>>,
-    #[serde(skip_serializing_if = "FederatedFacets::is_empty")]
-    pub facets_by_index: FederatedFacets,
-
-    // These fields are only used for analytics purposes
-    #[serde(skip)]
-    pub degraded: bool,
-    #[serde(skip)]
-    pub used_negative_operator: bool,
-}
-
-impl fmt::Debug for FederatedSearchResult {
-    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
-        let FederatedSearchResult {
-            hits,
-            processing_time_ms,
-            hits_info,
-            semantic_hit_count,
-            degraded,
-            used_negative_operator,
-            facet_distribution,
-            facet_stats,
-            facets_by_index,
-        } = self;
-
-        let mut debug = f.debug_struct("SearchResult");
-        // The most important thing when looking at a search result is the time it took to process
-        debug.field("processing_time_ms", &processing_time_ms);
-        debug.field("hits", &format!("[{} hits returned]", hits.len()));
-        debug.field("hits_info", &hits_info);
-        if *used_negative_operator {
-            debug.field("used_negative_operator", used_negative_operator);
-        }
-        if *degraded {
-            debug.field("degraded", degraded);
-        }
-        if let Some(facet_distribution) = facet_distribution {
-            debug.field("facet_distribution", &facet_distribution);
-        }
-        if let Some(facet_stats) = facet_stats {
-            debug.field("facet_stats", &facet_stats);
-        }
-        if let Some(semantic_hit_count) = semantic_hit_count {
-            debug.field("semantic_hit_count", &semantic_hit_count);
-        }
-        if !facets_by_index.is_empty() {
-            debug.field("facets_by_index", &facets_by_index);
-        }
-
-        debug.finish()
-    }
-}
-
-struct WeightedScore<'a> {
-    details: &'a [ScoreDetails],
-    weight: f64,
-}
-
-impl<'a> WeightedScore<'a> {
-    pub fn new(details: &'a [ScoreDetails], weight: f64) -> Self {
-        Self { details, weight }
-    }
-
-    pub fn weighted_global_score(&self) -> f64 {
-        ScoreDetails::global_score(self.details.iter()) * self.weight
-    }
-
-    pub fn compare_weighted_global_scores(&self, other: &Self) -> Ordering {
-        self.weighted_global_score()
-            .partial_cmp(&other.weighted_global_score())
-            // both are numbers, possibly infinite
-            .unwrap()
-    }
-
-    pub fn compare(&self, other: &Self) -> Ordering {
-        let mut left_it = ScoreDetails::score_values(self.details.iter());
-        let mut right_it = ScoreDetails::score_values(other.details.iter());
-
-        loop {
-            let left = left_it.next();
-            let right = right_it.next();
-
-            match (left, right) {
-                (None, None) => return Ordering::Equal,
-                (None, Some(_)) => return Ordering::Less,
-                (Some(_), None) => return Ordering::Greater,
-                (Some(ScoreValue::Score(left)), Some(ScoreValue::Score(right))) => {
-                    let left = left * self.weight;
-                    let right = right * other.weight;
-                    if (left - right).abs() <= f64::EPSILON {
-                        continue;
-                    }
-                    return left.partial_cmp(&right).unwrap();
-                }
-                (Some(ScoreValue::Sort(left)), Some(ScoreValue::Sort(right))) => {
-                    match left.partial_cmp(right) {
-                        Some(Ordering::Equal) => continue,
-                        Some(order) => return order,
-                        None => return self.compare_weighted_global_scores(other),
-                    }
-                }
-                (Some(ScoreValue::GeoSort(left)), Some(ScoreValue::GeoSort(right))) => {
-                    match left.partial_cmp(right) {
-                        Some(Ordering::Equal) => continue,
-                        Some(order) => return order,
-                        None => {
-                            return self.compare_weighted_global_scores(other);
-                        }
-                    }
-                }
-                // not comparable details, use global
-                (Some(ScoreValue::Score(_)), Some(_))
-                | (Some(_), Some(ScoreValue::Score(_)))
-                | (Some(ScoreValue::GeoSort(_)), Some(ScoreValue::Sort(_)))
-                | (Some(ScoreValue::Sort(_)), Some(ScoreValue::GeoSort(_))) => {
-                    let left_count = left_it.count();
-                    let right_count = right_it.count();
-                    // compare how many remaining groups of rules each side has.
-                    // the group with the most remaining groups wins.
-                    return left_count
-                        .cmp(&right_count)
-                        // breaks ties with the global ranking score
-                        .then_with(|| self.compare_weighted_global_scores(other));
-                }
-            }
-        }
-    }
-}
-
-struct QueryByIndex {
-    query: SearchQuery,
-    federation_options: FederationOptions,
-    query_index: usize,
-}
-
-struct SearchResultByQuery<'a> {
-    documents_ids: Vec<DocumentId>,
-    document_scores: Vec<Vec<ScoreDetails>>,
-    federation_options: FederationOptions,
-    hit_maker: HitMaker<'a>,
-    query_index: usize,
-}
-
-struct SearchResultByQueryIter<'a> {
-    it: Zip<IntoIter<DocumentId>, IntoIter<Vec<ScoreDetails>>>,
-    federation_options: FederationOptions,
-    hit_maker: Rc<HitMaker<'a>>,
-    query_index: usize,
-}
-
-impl<'a> SearchResultByQueryIter<'a> {
-    fn new(
-        SearchResultByQuery {
-            documents_ids,
-            document_scores,
-            federation_options,
-            hit_maker,
-            query_index,
-        }: SearchResultByQuery<'a>,
-    ) -> Self {
-        let it = documents_ids.into_iter().zip(document_scores);
-        Self { it, federation_options, hit_maker: Rc::new(hit_maker), query_index }
-    }
-}
-
-struct SearchResultByQueryIterItem<'a> {
-    docid: DocumentId,
-    score: Vec<ScoreDetails>,
-    federation_options: FederationOptions,
-    hit_maker: Rc<HitMaker<'a>>,
-    query_index: usize,
-}
-
-fn merge_index_local_results(
-    results_by_query: Vec<SearchResultByQuery<'_>>,
-) -> impl Iterator<Item = SearchResultByQueryIterItem> + '_ {
-    itertools::kmerge_by(
-        results_by_query.into_iter().map(SearchResultByQueryIter::new),
-        |left: &SearchResultByQueryIterItem, right: &SearchResultByQueryIterItem| {
-            let left_score = WeightedScore::new(&left.score, *left.federation_options.weight);
-            let right_score = WeightedScore::new(&right.score, *right.federation_options.weight);
-
-            match left_score.compare(&right_score) {
-                // the biggest score goes first
-                Ordering::Greater => true,
-                // break ties using query index
-                Ordering::Equal => left.query_index < right.query_index,
-                Ordering::Less => false,
-            }
-        },
-    )
-}
-
-fn merge_index_global_results(
-    results_by_index: Vec<SearchResultByIndex>,
-) -> impl Iterator<Item = SearchHitByIndex> {
-    itertools::kmerge_by(
-        results_by_index.into_iter().map(|result_by_index| result_by_index.hits.into_iter()),
-        |left: &SearchHitByIndex, right: &SearchHitByIndex| {
-            let left_score = WeightedScore::new(&left.score, *left.federation_options.weight);
-            let right_score = WeightedScore::new(&right.score, *right.federation_options.weight);
-
-            match left_score.compare(&right_score) {
-                // the biggest score goes first
-                Ordering::Greater => true,
-                // break ties using query index
-                Ordering::Equal => left.query_index < right.query_index,
-                Ordering::Less => false,
-            }
-        },
-    )
-}
-
-impl<'a> Iterator for SearchResultByQueryIter<'a> {
-    type Item = SearchResultByQueryIterItem<'a>;
-
-    fn next(&mut self) -> Option<Self::Item> {
-        let (docid, score) = self.it.next()?;
-        Some(SearchResultByQueryIterItem {
-            docid,
-            score,
-            federation_options: self.federation_options,
-            hit_maker: Rc::clone(&self.hit_maker),
-            query_index: self.query_index,
-        })
-    }
-}
-
-struct SearchHitByIndex {
-    hit: SearchHit,
-    score: Vec<ScoreDetails>,
-    federation_options: FederationOptions,
-    query_index: usize,
-}
-
-struct SearchResultByIndex {
-    index: String,
-    hits: Vec<SearchHitByIndex>,
-    estimated_total_hits: usize,
-    degraded: bool,
-    used_negative_operator: bool,
-    facets: Option<ComputedFacets>,
-}
-
-#[derive(Debug, Clone, Default, Serialize, ToSchema)]
-pub struct FederatedFacets(pub BTreeMap<String, ComputedFacets>);
-
-impl FederatedFacets {
-    pub fn insert(&mut self, index: String, facets: Option<ComputedFacets>) {
-        if let Some(facets) = facets {
-            self.0.insert(index, facets);
-        }
-    }
-
-    pub fn is_empty(&self) -> bool {
-        self.0.is_empty()
-    }
-
-    pub fn merge(
-        self,
-        MergeFacets { max_values_per_facet }: MergeFacets,
-        facet_order: BTreeMap<String, (String, OrderBy)>,
-    ) -> Option<ComputedFacets> {
-        if self.is_empty() {
-            return None;
-        }
-
-        let mut distribution: BTreeMap<String, _> = Default::default();
-        let mut stats: BTreeMap<String, FacetStats> = Default::default();
-
-        for facets_by_index in self.0.into_values() {
-            for (facet, index_distribution) in facets_by_index.distribution {
-                match distribution.entry(facet) {
-                    std::collections::btree_map::Entry::Vacant(entry) => {
-                        entry.insert(index_distribution);
-                    }
-                    std::collections::btree_map::Entry::Occupied(mut entry) => {
-                        let distribution = entry.get_mut();
-
-                        for (value, index_count) in index_distribution {
-                            distribution
-                                .entry(value)
-                                .and_modify(|count| *count += index_count)
-                                .or_insert(index_count);
-                        }
-                    }
-                }
-            }
-
-            for (facet, index_stats) in facets_by_index.stats {
-                match stats.entry(facet) {
-                    std::collections::btree_map::Entry::Vacant(entry) => {
-                        entry.insert(index_stats);
-                    }
-                    std::collections::btree_map::Entry::Occupied(mut entry) => {
-                        let stats = entry.get_mut();
-
-                        stats.min = f64::min(stats.min, index_stats.min);
-                        stats.max = f64::max(stats.max, index_stats.max);
-                    }
-                }
-            }
-        }
-
-        // fixup order
-        for (facet, values) in &mut distribution {
-            let order_by = facet_order.get(facet).map(|(_, order)| *order).unwrap_or_default();
-
-            match order_by {
-                OrderBy::Lexicographic => {
-                    values.sort_unstable_by(|left, _, right, _| left.cmp(right))
-                }
-                OrderBy::Count => {
-                    values.sort_unstable_by(|_, left, _, right| {
-                        left.cmp(right)
-                            // biggest first
-                            .reverse()
-                    })
-                }
-            }
-
-            if let Some(max_values_per_facet) = max_values_per_facet {
-                values.truncate(max_values_per_facet)
-            };
-        }
-
-        Some(ComputedFacets { distribution, stats })
-    }
-}
-
-pub fn perform_federated_search(
-    index_scheduler: &IndexScheduler,
-    queries: Vec<SearchQueryWithIndex>,
-    mut federation: Federation,
-    features: RoFeatures,
-) -> Result<FederatedSearchResult, ResponseError> {
-    let before_search = std::time::Instant::now();
-
-    // this implementation partition the queries by index to guarantee an important property:
-    // - all the queries to a particular index use the same read transaction.
-    // This is an important property, otherwise we cannot guarantee the self-consistency of the results.
-
-    // 1. partition queries by index
-    let mut queries_by_index: BTreeMap<String, Vec<QueryByIndex>> = Default::default();
-    for (query_index, federated_query) in queries.into_iter().enumerate() {
-        if let Some(pagination_field) = federated_query.has_pagination() {
-            return Err(MeilisearchHttpError::PaginationInFederatedQuery(
-                query_index,
-                pagination_field,
-            )
-            .into());
-        }
-
-        if let Some(facets) = federated_query.has_facets() {
-            let facets = facets.to_owned();
-            return Err(MeilisearchHttpError::FacetsInFederatedQuery(
-                query_index,
-                federated_query.index_uid.into_inner(),
-                facets,
-            )
-            .into());
-        }
-
-        let (index_uid, query, federation_options) = federated_query.into_index_query_federation();
-
-        queries_by_index.entry(index_uid.into_inner()).or_default().push(QueryByIndex {
-            query,
-            federation_options: federation_options.unwrap_or_default(),
-            query_index,
-        })
-    }
-
-    // 2. perform queries, merge and make hits index by index
-    let required_hit_count = federation.limit + federation.offset;
-
-    // In step (2), semantic_hit_count will be set to Some(0) if any search kind uses semantic
-    // Then in step (3), we'll update its value if there is any semantic search
-    let mut semantic_hit_count = None;
-    let mut results_by_index = Vec::with_capacity(queries_by_index.len());
-    let mut previous_query_data: Option<(RankingRules, usize, String)> = None;
-
-    // remember the order and name of first index for each facet when merging with index settings
-    // to detect if the order is inconsistent for a facet.
-    let mut facet_order: Option<BTreeMap<String, (String, OrderBy)>> = match federation.merge_facets
-    {
-        Some(MergeFacets { .. }) => Some(Default::default()),
-        _ => None,
-    };
-
-    for (index_uid, queries) in queries_by_index {
-        let first_query_index = queries.first().map(|query| query.query_index);
-
-        let index = match index_scheduler.index(&index_uid) {
-            Ok(index) => index,
-            Err(err) => {
-                let mut err = ResponseError::from(err);
-                // Patch the HTTP status code to 400 as it defaults to 404 for `index_not_found`, but
-                // here the resource not found is not part of the URL.
-                err.code = StatusCode::BAD_REQUEST;
-                if let Some(query_index) = first_query_index {
-                    err.message = format!("Inside `.queries[{}]`: {}", query_index, err.message);
-                }
-                return Err(err);
-            }
-        };
-
-        // Important: this is the only transaction we'll use for this index during this federated search
-        let rtxn = index.read_txn()?;
-
-        let criteria = index.criteria(&rtxn)?;
-
-        let dictionary = index.dictionary(&rtxn)?;
-        let dictionary: Option<Vec<_>> =
-            dictionary.as_ref().map(|x| x.iter().map(String::as_str).collect());
-        let separators = index.allowed_separators(&rtxn)?;
-        let separators: Option<Vec<_>> =
-            separators.as_ref().map(|x| x.iter().map(String::as_str).collect());
-
-        // each query gets its individual cutoff
-        let cutoff = index.search_cutoff(&rtxn)?;
-
-        let mut degraded = false;
-        let mut used_negative_operator = false;
-        let mut candidates = RoaringBitmap::new();
-
-        let facets_by_index = federation.facets_by_index.remove(&index_uid).flatten();
-
-        // TODO: recover the max size + facets_by_index as return value of this function so as not to ask it for all queries
-        if let Err(mut error) =
-            check_facet_order(&mut facet_order, &index_uid, &facets_by_index, &index, &rtxn)
-        {
-            error.message = format!(
-                "Inside `.federation.facetsByIndex.{index_uid}`: {error}{}",
-                if let Some(query_index) = first_query_index {
-                    format!("\n - Note: index `{index_uid}` used in `.queries[{query_index}]`")
-                } else {
-                    Default::default()
-                }
-            );
-            return Err(error);
-        }
-
-        // 2.1. Compute all candidates for each query in the index
-        let mut results_by_query = Vec::with_capacity(queries.len());
-
-        for QueryByIndex { query, federation_options, query_index } in queries {
-            // use an immediately invoked lambda to capture the result without returning from the function
-
-            let res: Result<(), ResponseError> = (|| {
-                let search_kind =
-                    search_kind(&query, index_scheduler, index_uid.to_string(), &index)?;
-
-                let canonicalization_kind = match (&search_kind, &query.q) {
-                    (SearchKind::SemanticOnly { .. }, _) => {
-                        ranking_rules::CanonicalizationKind::Vector
-                    }
-                    (_, Some(q)) if !q.is_empty() => ranking_rules::CanonicalizationKind::Keyword,
-                    _ => ranking_rules::CanonicalizationKind::Placeholder,
-                };
-
-                let sort = if let Some(sort) = &query.sort {
-                    let sorts: Vec<_> =
-                        match sort.iter().map(|s| milli::AscDesc::from_str(s)).collect() {
-                            Ok(sorts) => sorts,
-                            Err(asc_desc_error) => {
-                                return Err(milli::Error::from(milli::SortError::from(
-                                    asc_desc_error,
-                                ))
-                                .into())
-                            }
-                        };
-                    Some(sorts)
-                } else {
-                    None
-                };
-
-                let ranking_rules = ranking_rules::RankingRules::new(
-                    criteria.clone(),
-                    sort,
-                    query.matching_strategy.into(),
-                    canonicalization_kind,
-                );
-
-                if let Some((previous_ranking_rules, previous_query_index, previous_index_uid)) =
-                    previous_query_data.take()
-                {
-                    if let Err(error) = ranking_rules.is_compatible_with(&previous_ranking_rules) {
-                        return Err(error.to_response_error(
-                            &ranking_rules,
-                            &previous_ranking_rules,
-                            query_index,
-                            previous_query_index,
-                            &index_uid,
-                            &previous_index_uid,
-                        ));
-                    }
-                    previous_query_data = if previous_ranking_rules.constraint_count()
-                        > ranking_rules.constraint_count()
-                    {
-                        Some((previous_ranking_rules, previous_query_index, previous_index_uid))
-                    } else {
-                        Some((ranking_rules, query_index, index_uid.clone()))
-                    };
-                } else {
-                    previous_query_data = Some((ranking_rules, query_index, index_uid.clone()));
-                }
-
-                match search_kind {
-                    SearchKind::KeywordOnly => {}
-                    _ => semantic_hit_count = Some(0),
-                }
-
-                let retrieve_vectors = RetrieveVectors::new(query.retrieve_vectors);
-
-                let time_budget = match cutoff {
-                    Some(cutoff) => TimeBudget::new(Duration::from_millis(cutoff)),
-                    None => TimeBudget::default(),
-                };
-
-                let (mut search, _is_finite_pagination, _max_total_hits, _offset) =
-                    prepare_search(&index, &rtxn, &query, &search_kind, time_budget, features)?;
-
-                search.scoring_strategy(milli::score_details::ScoringStrategy::Detailed);
-                search.offset(0);
-                search.limit(required_hit_count);
-
-                let (result, _semantic_hit_count) =
-                    super::search_from_kind(index_uid.to_string(), search_kind, search)?;
-                let format = AttributesFormat {
-                    attributes_to_retrieve: query.attributes_to_retrieve,
-                    retrieve_vectors,
-                    attributes_to_highlight: query.attributes_to_highlight,
-                    attributes_to_crop: query.attributes_to_crop,
-                    crop_length: query.crop_length,
-                    crop_marker: query.crop_marker,
-                    highlight_pre_tag: query.highlight_pre_tag,
-                    highlight_post_tag: query.highlight_post_tag,
-                    show_matches_position: query.show_matches_position,
-                    sort: query.sort,
-                    show_ranking_score: query.show_ranking_score,
-                    show_ranking_score_details: query.show_ranking_score_details,
-                    locales: query.locales.map(|l| l.iter().copied().map(Into::into).collect()),
-                };
-
-                let milli::SearchResult {
-                    matching_words,
-                    candidates: query_candidates,
-                    documents_ids,
-                    document_scores,
-                    degraded: query_degraded,
-                    used_negative_operator: query_used_negative_operator,
-                } = result;
-
-                candidates |= query_candidates;
-                degraded |= query_degraded;
-                used_negative_operator |= query_used_negative_operator;
-
-                let tokenizer = HitMaker::tokenizer(dictionary.as_deref(), separators.as_deref());
-
-                let formatter_builder = HitMaker::formatter_builder(matching_words, tokenizer);
-
-                let hit_maker =
-                    HitMaker::new(&index, &rtxn, format, formatter_builder).map_err(|e| {
-                        MeilisearchHttpError::from_milli(e, Some(index_uid.to_string()))
-                    })?;
-
-                results_by_query.push(SearchResultByQuery {
-                    federation_options,
-                    hit_maker,
-                    query_index,
-                    documents_ids,
-                    document_scores,
-                });
-                Ok(())
-            })();
-
-            if let Err(mut error) = res {
-                error.message = format!("Inside `.queries[{query_index}]`: {}", error.message);
-                return Err(error);
-            }
-        }
-        // 2.2. merge inside index
-        let mut documents_seen = RoaringBitmap::new();
-        let merged_result: Result<Vec<_>, ResponseError> =
-            merge_index_local_results(results_by_query)
-                // skip documents we've already seen & mark that we saw the current document
-                .filter(|SearchResultByQueryIterItem { docid, .. }| documents_seen.insert(*docid))
-                .take(required_hit_count)
-                // 2.3 make hits
-                .map(
-                    |SearchResultByQueryIterItem {
-                         docid,
-                         score,
-                         federation_options,
-                         hit_maker,
-                         query_index,
-                     }| {
-                        let mut hit = hit_maker.make_hit(docid, &score)?;
-                        let weighted_score =
-                            ScoreDetails::global_score(score.iter()) * (*federation_options.weight);
-
-                        let _federation = serde_json::json!(
-                            {
-                                "indexUid": index_uid,
-                                "queriesPosition": query_index,
-                                "weightedRankingScore": weighted_score,
-                            }
-                        );
-                        hit.document.insert("_federation".to_string(), _federation);
-                        Ok(SearchHitByIndex { hit, score, federation_options, query_index })
-                    },
-                )
-                .collect();
-
-        let merged_result = merged_result?;
-
-        let estimated_total_hits = candidates.len() as usize;
-
-        let facets = facets_by_index
-            .map(|facets_by_index| {
-                compute_facet_distribution_stats(
-                    &facets_by_index,
-                    &index,
-                    &rtxn,
-                    candidates,
-                    super::Route::MultiSearch,
-                )
-            })
-            .transpose()
-            .map_err(|mut error| {
-                error.message = format!(
-                    "Inside `.federation.facetsByIndex.{index_uid}`: {}{}",
-                    error.message,
-                    if let Some(query_index) = first_query_index {
-                        format!("\n - Note: index `{index_uid}` used in `.queries[{query_index}]`")
-                    } else {
-                        Default::default()
-                    }
-                );
-                error
-            })?;
-
-        results_by_index.push(SearchResultByIndex {
-            index: index_uid,
-            hits: merged_result,
-            estimated_total_hits,
-            degraded,
-            used_negative_operator,
-            facets,
-        });
-    }
-
-    // bonus step, make sure to return an error if an index wants a non-faceted field, even if no query actually uses that index.
-    for (index_uid, facets) in federation.facets_by_index {
-        let index = match index_scheduler.index(&index_uid) {
-            Ok(index) => index,
-            Err(err) => {
-                let mut err = ResponseError::from(err);
-                // Patch the HTTP status code to 400 as it defaults to 404 for `index_not_found`, but
-                // here the resource not found is not part of the URL.
-                err.code = StatusCode::BAD_REQUEST;
-                err.message = format!(
-                    "Inside `.federation.facetsByIndex.{index_uid}`: {}\n - Note: index `{index_uid}` is not used in queries",
-                    err.message
-                );
-                return Err(err);
-            }
-        };
-
-        // Important: this is the only transaction we'll use for this index during this federated search
-        let rtxn = index.read_txn()?;
-
-        if let Err(mut error) =
-            check_facet_order(&mut facet_order, &index_uid, &facets, &index, &rtxn)
-        {
-            error.message = format!(
-                "Inside `.federation.facetsByIndex.{index_uid}`: {error}\n - Note: index `{index_uid}` is not used in queries",
-            );
-            return Err(error);
-        }
-
-        if let Some(facets) = facets {
-            if let Err(mut error) = compute_facet_distribution_stats(
-                &facets,
-                &index,
-                &rtxn,
-                Default::default(),
-                super::Route::MultiSearch,
-            ) {
-                error.message =
-                    format!("Inside `.federation.facetsByIndex.{index_uid}`: {}\n - Note: index `{index_uid}` is not used in queries", error.message);
-                return Err(error);
-            }
-        }
-    }
-
-    // 3. merge hits and metadata across indexes
-    // 3.1 merge metadata
-    let (estimated_total_hits, degraded, used_negative_operator, facets) = {
-        let mut estimated_total_hits = 0;
-        let mut degraded = false;
-        let mut used_negative_operator = false;
-
-        let mut facets: FederatedFacets = FederatedFacets::default();
-
-        for SearchResultByIndex {
-            index,
-            hits: _,
-            estimated_total_hits: estimated_total_hits_by_index,
-            facets: facets_by_index,
-            degraded: degraded_by_index,
-            used_negative_operator: used_negative_operator_by_index,
-        } in &mut results_by_index
-        {
-            estimated_total_hits += *estimated_total_hits_by_index;
-            degraded |= *degraded_by_index;
-            used_negative_operator |= *used_negative_operator_by_index;
-
-            let facets_by_index = std::mem::take(facets_by_index);
-            let index = std::mem::take(index);
-
-            facets.insert(index, facets_by_index);
-        }
-
-        (estimated_total_hits, degraded, used_negative_operator, facets)
-    };
-
-    // 3.2 merge hits
-    let merged_hits: Vec<_> = merge_index_global_results(results_by_index)
-        .skip(federation.offset)
-        .take(federation.limit)
-        .inspect(|hit| {
-            if let Some(semantic_hit_count) = &mut semantic_hit_count {
-                if hit.score.iter().any(|score| matches!(&score, ScoreDetails::Vector(_))) {
-                    *semantic_hit_count += 1;
-                }
-            }
-        })
-        .map(|hit| hit.hit)
-        .collect();
-
-    let (facet_distribution, facet_stats, facets_by_index) =
-        match federation.merge_facets.zip(facet_order) {
-            Some((merge_facets, facet_order)) => {
-                let facets = facets.merge(merge_facets, facet_order);
-
-                let (facet_distribution, facet_stats) = facets
-                    .map(|ComputedFacets { distribution, stats }| (distribution, stats))
-                    .unzip();
-
-                (facet_distribution, facet_stats, FederatedFacets::default())
-            }
-            None => (None, None, facets),
-        };
-
-    let search_result = FederatedSearchResult {
-        hits: merged_hits,
-        processing_time_ms: before_search.elapsed().as_millis(),
-        hits_info: HitsInfo::OffsetLimit {
-            limit: federation.limit,
-            offset: federation.offset,
-            estimated_total_hits,
-        },
-        semantic_hit_count,
-        degraded,
-        used_negative_operator,
-        facet_distribution,
-        facet_stats,
-        facets_by_index,
-    };
-
-    Ok(search_result)
-}
-
-fn check_facet_order(
-    facet_order: &mut Option<BTreeMap<String, (String, OrderBy)>>,
-    current_index: &str,
-    facets_by_index: &Option<Vec<String>>,
-    index: &milli::Index,
-    rtxn: &milli::heed::RoTxn<'_>,
-) -> Result<(), ResponseError> {
-    if let (Some(facet_order), Some(facets_by_index)) = (facet_order, facets_by_index) {
-        let index_facet_order = index.sort_facet_values_by(rtxn)?;
-        for facet in facets_by_index {
-            let index_facet_order = index_facet_order.get(facet);
-            let (previous_index, previous_facet_order) = facet_order
-                .entry(facet.to_owned())
-                .or_insert_with(|| (current_index.to_owned(), index_facet_order));
-            if previous_facet_order != &index_facet_order {
-                return Err(MeilisearchHttpError::InconsistentFacetOrder {
-                    facet: facet.clone(),
-                    previous_facet_order: *previous_facet_order,
-                    previous_uid: previous_index.clone(),
-                    current_uid: current_index.to_owned(),
-                    index_facet_order,
-                }
-                .into());
-            }
-        }
-    };
-    Ok(())
-}

crates/meilisearch/src/search/federated/mod.rs

@@ -0,0 +1,10 @@
+mod perform;
+mod proxy;
+mod types;
+mod weighted_scores;
+
+pub use perform::perform_federated_search;
+pub use proxy::{PROXY_SEARCH_HEADER, PROXY_SEARCH_HEADER_VALUE};
+pub use types::{
+    FederatedSearch, FederatedSearchResult, Federation, FederationOptions, MergeFacets,
+};

crates/meilisearch/src/search/federated/proxy.rs

@@ -0,0 +1,267 @@
+pub use error::ProxySearchError;
+use error::ReqwestErrorWithoutUrl;
+use meilisearch_types::features::Remote;
+use rand::Rng as _;
+use reqwest::{Client, Response, StatusCode};
+use serde::de::DeserializeOwned;
+use serde_json::Value;
+
+use super::types::{FederatedSearch, FederatedSearchResult, Federation};
+use crate::search::SearchQueryWithIndex;
+
+pub const PROXY_SEARCH_HEADER: &str = "Meili-Proxy-Search";
+pub const PROXY_SEARCH_HEADER_VALUE: &str = "true";
+
+mod error {
+    use meilisearch_types::error::ResponseError;
+    use reqwest::StatusCode;
+
+    #[derive(Debug, thiserror::Error)]
+    pub enum ProxySearchError {
+        #[error("{0}")]
+        CouldNotSendRequest(ReqwestErrorWithoutUrl),
+        #[error("could not authenticate against the remote host\n  - hint: check that the remote instance was registered with a valid API key having the `search` action")]
+        AuthenticationError,
+        #[error(
+            "could not parse response from the remote host as a federated search response{}\n  - hint: check that the remote instance is a Meilisearch instance running the same version",
+            response_from_remote(response)
+        )]
+        CouldNotParseResponse { response: Result<String, ReqwestErrorWithoutUrl> },
+        #[error("remote host responded with code {}{}\n  - hint: check that the remote instance has the correct index configuration for that request\n  - hint: check that the `network` experimental feature is enabled on the remote instance", status_code.as_u16(), response_from_remote(response))]
+        BadRequest { status_code: StatusCode, response: Result<String, ReqwestErrorWithoutUrl> },
+        #[error("remote host did not answer before the deadline")]
+        Timeout,
+        #[error("remote hit does not contain `{0}`\n  - hint: check that the remote instance is a Meilisearch instance running the same version")]
+        MissingPathInResponse(&'static str),
+        #[error("remote host responded with code {}{}", status_code.as_u16(), response_from_remote(response))]
+        RemoteError { status_code: StatusCode, response: Result<String, ReqwestErrorWithoutUrl> },
+        #[error("remote hit contains an unexpected value at path `{path}`: expected {expected_type}, received `{received_value}`\n  - hint: check that the remote instance is a Meilisearch instance running the same version")]
+        UnexpectedValueInPath {
+            path: &'static str,
+            expected_type: &'static str,
+            received_value: String,
+        },
+        #[error("could not parse weighted score values in the remote hit: {0}")]
+        CouldNotParseWeightedScoreValues(serde_json::Error),
+    }
+
+    impl ProxySearchError {
+        pub fn as_response_error(&self) -> ResponseError {
+            use meilisearch_types::error::Code;
+            let message = self.to_string();
+            let code = match self {
+                ProxySearchError::CouldNotSendRequest(_) => Code::RemoteCouldNotSendRequest,
+                ProxySearchError::AuthenticationError => Code::RemoteInvalidApiKey,
+                ProxySearchError::BadRequest { .. } => Code::RemoteBadRequest,
+                ProxySearchError::Timeout => Code::RemoteTimeout,
+                ProxySearchError::RemoteError { .. } => Code::RemoteRemoteError,
+                ProxySearchError::CouldNotParseResponse { .. }
+                | ProxySearchError::MissingPathInResponse(_)
+                | ProxySearchError::UnexpectedValueInPath { .. }
+                | ProxySearchError::CouldNotParseWeightedScoreValues(_) => Code::RemoteBadResponse,
+            };
+            ResponseError::from_msg(message, code)
+        }
+    }
+
+    #[derive(Debug, thiserror::Error)]
+    #[error(transparent)]
+    pub struct ReqwestErrorWithoutUrl(reqwest::Error);
+    impl ReqwestErrorWithoutUrl {
+        pub fn new(inner: reqwest::Error) -> Self {
+            Self(inner.without_url())
+        }
+    }
+
+    fn response_from_remote(response: &Result<String, ReqwestErrorWithoutUrl>) -> String {
+        match response {
+            Ok(response) => {
+                format!(":\n  - response from remote: {}", response)
+            }
+            Err(error) => {
+                format!(":\n  - additionally, could not retrieve response from remote: {error}")
+            }
+        }
+    }
+}
+
+#[derive(Clone)]
+pub struct ProxySearchParams {
+    pub deadline: Option<std::time::Instant>,
+    pub try_count: u32,
+    pub client: reqwest::Client,
+}
+
+/// Performs a federated search on a remote host and returns the results
+pub async fn proxy_search(
+    node: &Remote,
+    queries: Vec<SearchQueryWithIndex>,
+    federation: Federation,
+    params: &ProxySearchParams,
+) -> Result<FederatedSearchResult, ProxySearchError> {
+    let url = format!("{}/multi-search", node.url);
+
+    let federated = FederatedSearch { queries, federation: Some(federation) };
+
+    let search_api_key = node.search_api_key.as_deref();
+
+    let max_deadline = std::time::Instant::now() + std::time::Duration::from_secs(5);
+
+    let deadline = if let Some(deadline) = params.deadline {
+        std::time::Instant::min(deadline, max_deadline)
+    } else {
+        max_deadline
+    };
+
+    for i in 0..params.try_count {
+        match try_proxy_search(&url, search_api_key, &federated, &params.client, deadline).await {
+            Ok(response) => return Ok(response),
+            Err(retry) => {
+                let duration = retry.into_duration(i)?;
+                tokio::time::sleep(duration).await;
+            }
+        }
+    }
+    try_proxy_search(&url, search_api_key, &federated, &params.client, deadline)
+        .await
+        .map_err(Retry::into_error)
+}
+
+async fn try_proxy_search(
+    url: &str,
+    search_api_key: Option<&str>,
+    federated: &FederatedSearch,
+    client: &Client,
+    deadline: std::time::Instant,
+) -> Result<FederatedSearchResult, Retry> {
+    let timeout = deadline.saturating_duration_since(std::time::Instant::now());
+
+    let request = client.post(url).json(&federated).timeout(timeout);
+    let request = if let Some(search_api_key) = search_api_key {
+        request.bearer_auth(search_api_key)
+    } else {
+        request
+    };
+    let request = request.header(PROXY_SEARCH_HEADER, PROXY_SEARCH_HEADER_VALUE);
+
+    let response = request.send().await;
+    let response = match response {
+        Ok(response) => response,
+        Err(error) if error.is_timeout() => return Err(Retry::give_up(ProxySearchError::Timeout)),
+        Err(error) => {
+            return Err(Retry::retry_later(ProxySearchError::CouldNotSendRequest(
+                ReqwestErrorWithoutUrl::new(error),
+            )))
+        }
+    };
+
+    match response.status() {
+        status_code if status_code.is_success() => (),
+        StatusCode::UNAUTHORIZED | StatusCode::FORBIDDEN => {
+            return Err(Retry::give_up(ProxySearchError::AuthenticationError))
+        }
+        status_code if status_code.is_client_error() => {
+            let response = parse_error(response).await;
+            return Err(Retry::give_up(ProxySearchError::BadRequest { status_code, response }));
+        }
+        status_code if status_code.is_server_error() => {
+            let response = parse_error(response).await;
+            return Err(Retry::retry_later(ProxySearchError::RemoteError {
+                status_code,
+                response,
+            }));
+        }
+        status_code => {
+            tracing::warn!(
+                status_code = status_code.as_u16(),
+                "remote replied with unexpected status code"
+            );
+        }
+    }
+
+    let response = match parse_response(response).await {
+        Ok(response) => response,
+        Err(response) => {
+            return Err(Retry::retry_later(ProxySearchError::CouldNotParseResponse { response }))
+        }
+    };
+
+    Ok(response)
+}
+
+/// Always parse the body of the response of a failed request as JSON.
+async fn parse_error(response: Response) -> Result<String, ReqwestErrorWithoutUrl> {
+    let bytes = match response.bytes().await {
+        Ok(bytes) => bytes,
+        Err(error) => return Err(ReqwestErrorWithoutUrl::new(error)),
+    };
+
+    Ok(parse_bytes_as_error(&bytes))
+}
+
+fn parse_bytes_as_error(bytes: &[u8]) -> String {
+    match serde_json::from_slice::<Value>(bytes) {
+        Ok(value) => value.to_string(),
+        Err(_) => String::from_utf8_lossy(bytes).into_owned(),
+    }
+}
+
+async fn parse_response<T: DeserializeOwned>(
+    response: Response,
+) -> Result<T, Result<String, ReqwestErrorWithoutUrl>> {
+    let bytes = match response.bytes().await {
+        Ok(bytes) => bytes,
+        Err(error) => return Err(Err(ReqwestErrorWithoutUrl::new(error))),
+    };
+
+    match serde_json::from_slice::<T>(&bytes) {
+        Ok(value) => Ok(value),
+        Err(_) => Err(Ok(parse_bytes_as_error(&bytes))),
+    }
+}
+
+pub struct Retry {
+    error: ProxySearchError,
+    strategy: RetryStrategy,
+}
+
+pub enum RetryStrategy {
+    GiveUp,
+    Retry,
+}
+
+impl Retry {
+    pub fn give_up(error: ProxySearchError) -> Self {
+        Self { error, strategy: RetryStrategy::GiveUp }
+    }
+
+    pub fn retry_later(error: ProxySearchError) -> Self {
+        Self { error, strategy: RetryStrategy::Retry }
+    }
+
+    pub fn into_duration(self, attempt: u32) -> Result<std::time::Duration, ProxySearchError> {
+        match self.strategy {
+            RetryStrategy::GiveUp => Err(self.error),
+            RetryStrategy::Retry => {
+                let retry_duration = std::time::Duration::from_nanos((10u64).pow(attempt));
+                let retry_duration = retry_duration.min(std::time::Duration::from_millis(100)); // don't wait more than 100ms
+
+                // randomly up to double the retry duration
+                let retry_duration = retry_duration
+                    + rand::thread_rng().gen_range(std::time::Duration::ZERO..retry_duration);
+
+                tracing::warn!(
+                    "Attempt #{}, failed with {}, retrying after {}ms.",
+                    attempt,
+                    self.error,
+                    retry_duration.as_millis()
+                );
+                Ok(retry_duration)
+            }
+        }
+    }
+
+    pub fn into_error(self) -> ProxySearchError {
+        self.error
+    }
+}

crates/meilisearch/src/search/federated/types.rs

@@ -0,0 +1,322 @@
+use std::collections::btree_map::Entry;
+use std::collections::BTreeMap;
+use std::fmt;
+use std::vec::Vec;
+
+use indexmap::IndexMap;
+use meilisearch_types::deserr::DeserrJsonError;
+use meilisearch_types::error::deserr_codes::{
+    InvalidMultiSearchFacetsByIndex, InvalidMultiSearchMaxValuesPerFacet,
+    InvalidMultiSearchMergeFacets, InvalidMultiSearchQueryPosition, InvalidMultiSearchRemote,
+    InvalidMultiSearchWeight, InvalidSearchLimit, InvalidSearchOffset,
+};
+use meilisearch_types::error::ResponseError;
+use meilisearch_types::index_uid::IndexUid;
+use meilisearch_types::milli::order_by_map::OrderByMap;
+use meilisearch_types::milli::OrderBy;
+use serde::{Deserialize, Serialize};
+use utoipa::ToSchema;
+
+use super::super::{ComputedFacets, FacetStats, HitsInfo, SearchHit, SearchQueryWithIndex};
+
+pub const DEFAULT_FEDERATED_WEIGHT: f64 = 1.0;
+
+// fields in the response
+pub const FEDERATION_HIT: &str = "_federation";
+pub const INDEX_UID: &str = "indexUid";
+pub const QUERIES_POSITION: &str = "queriesPosition";
+pub const WEIGHTED_RANKING_SCORE: &str = "weightedRankingScore";
+pub const WEIGHTED_SCORE_VALUES: &str = "weightedScoreValues";
+pub const FEDERATION_REMOTE: &str = "remote";
+
+#[derive(Debug, Default, Clone, PartialEq, Serialize, deserr::Deserr, ToSchema)]
+#[deserr(error = DeserrJsonError, rename_all = camelCase, deny_unknown_fields)]
+#[serde(rename_all = "camelCase")]
+
+pub struct FederationOptions {
+    #[deserr(default, error = DeserrJsonError<InvalidMultiSearchWeight>)]
+    #[schema(value_type = f64)]
+    pub weight: Weight,
+
+    #[deserr(default, error = DeserrJsonError<InvalidMultiSearchRemote>)]
+    pub remote: Option<String>,
+
+    #[deserr(default, error = DeserrJsonError<InvalidMultiSearchQueryPosition>)]
+    pub query_position: Option<usize>,
+}
+
+#[derive(Debug, Clone, Copy, PartialEq, Serialize, deserr::Deserr)]
+#[deserr(try_from(f64) = TryFrom::try_from -> InvalidMultiSearchWeight)]
+pub struct Weight(f64);
+
+impl Default for Weight {
+    fn default() -> Self {
+        Weight(DEFAULT_FEDERATED_WEIGHT)
+    }
+}
+
+impl std::convert::TryFrom<f64> for Weight {
+    type Error = InvalidMultiSearchWeight;
+
+    fn try_from(f: f64) -> Result<Self, Self::Error> {
+        if f < 0.0 {
+            Err(InvalidMultiSearchWeight)
+        } else {
+            Ok(Weight(f))
+        }
+    }
+}
+
+impl std::ops::Deref for Weight {
+    type Target = f64;
+
+    fn deref(&self) -> &Self::Target {
+        &self.0
+    }
+}
+
+#[derive(Debug, Clone, deserr::Deserr, Serialize, ToSchema)]
+#[deserr(error = DeserrJsonError, rename_all = camelCase, deny_unknown_fields)]
+#[schema(rename_all = "camelCase")]
+#[serde(rename_all = "camelCase")]
+pub struct Federation {
+    #[deserr(default = super::super::DEFAULT_SEARCH_LIMIT(), error = DeserrJsonError<InvalidSearchLimit>)]
+    pub limit: usize,
+    #[deserr(default = super::super::DEFAULT_SEARCH_OFFSET(), error = DeserrJsonError<InvalidSearchOffset>)]
+    pub offset: usize,
+    #[deserr(default, error = DeserrJsonError<InvalidMultiSearchFacetsByIndex>)]
+    pub facets_by_index: BTreeMap<IndexUid, Option<Vec<String>>>,
+    #[deserr(default, error = DeserrJsonError<InvalidMultiSearchMergeFacets>)]
+    pub merge_facets: Option<MergeFacets>,
+}
+
+#[derive(Copy, Clone, Debug, deserr::Deserr, Serialize, Default, ToSchema)]
+#[deserr(error = DeserrJsonError<InvalidMultiSearchMergeFacets>, rename_all = camelCase, deny_unknown_fields)]
+#[schema(rename_all = "camelCase")]
+#[serde(rename_all = "camelCase")]
+pub struct MergeFacets {
+    #[deserr(default, error = DeserrJsonError<InvalidMultiSearchMaxValuesPerFacet>)]
+    pub max_values_per_facet: Option<usize>,
+}
+
+#[derive(Debug, deserr::Deserr, Serialize, ToSchema)]
+#[deserr(error = DeserrJsonError, rename_all = camelCase, deny_unknown_fields)]
+#[schema(rename_all = "camelCase")]
+#[serde(rename_all = "camelCase")]
+pub struct FederatedSearch {
+    pub queries: Vec<SearchQueryWithIndex>,
+    #[deserr(default)]
+    pub federation: Option<Federation>,
+}
+
+#[derive(Serialize, Deserialize, Clone, ToSchema)]
+#[serde(rename_all = "camelCase")]
+#[schema(rename_all = "camelCase")]
+pub struct FederatedSearchResult {
+    pub hits: Vec<SearchHit>,
+    pub processing_time_ms: u128,
+    #[serde(flatten)]
+    pub hits_info: HitsInfo,
+
+    #[serde(default, skip_serializing_if = "Option::is_none")]
+    pub semantic_hit_count: Option<u32>,
+
+    #[serde(default, skip_serializing_if = "Option::is_none")]
+    #[schema(value_type = Option<BTreeMap<String, BTreeMap<String, u64>>>)]
+    pub facet_distribution: Option<BTreeMap<String, IndexMap<String, u64>>>,
+    #[serde(default, skip_serializing_if = "Option::is_none")]
+    pub facet_stats: Option<BTreeMap<String, FacetStats>>,
+    #[serde(default, skip_serializing_if = "FederatedFacets::is_empty")]
+    pub facets_by_index: FederatedFacets,
+
+    #[serde(default, skip_serializing_if = "Option::is_none")]
+    pub remote_errors: Option<BTreeMap<String, ResponseError>>,
+
+    // These fields are only used for analytics purposes
+    #[serde(skip)]
+    pub degraded: bool,
+    #[serde(skip)]
+    pub used_negative_operator: bool,
+}
+
+impl fmt::Debug for FederatedSearchResult {
+    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+        let FederatedSearchResult {
+            hits,
+            processing_time_ms,
+            hits_info,
+            semantic_hit_count,
+            degraded,
+            used_negative_operator,
+            facet_distribution,
+            facet_stats,
+            facets_by_index,
+            remote_errors,
+        } = self;
+
+        let mut debug = f.debug_struct("SearchResult");
+        // The most important thing when looking at a search result is the time it took to process
+        debug.field("processing_time_ms", &processing_time_ms);
+        debug.field("hits", &format!("[{} hits returned]", hits.len()));
+        debug.field("hits_info", &hits_info);
+        if *used_negative_operator {
+            debug.field("used_negative_operator", used_negative_operator);
+        }
+        if *degraded {
+            debug.field("degraded", degraded);
+        }
+        if let Some(facet_distribution) = facet_distribution {
+            debug.field("facet_distribution", &facet_distribution);
+        }
+        if let Some(facet_stats) = facet_stats {
+            debug.field("facet_stats", &facet_stats);
+        }
+        if let Some(semantic_hit_count) = semantic_hit_count {
+            debug.field("semantic_hit_count", &semantic_hit_count);
+        }
+        if !facets_by_index.is_empty() {
+            debug.field("facets_by_index", &facets_by_index);
+        }
+        if let Some(remote_errors) = remote_errors {
+            debug.field("remote_errors", &remote_errors);
+        }
+
+        debug.finish()
+    }
+}
+
+#[derive(Debug, Clone, Default, Serialize, Deserialize, ToSchema)]
+pub struct FederatedFacets(pub BTreeMap<String, ComputedFacets>);
+
+impl FederatedFacets {
+    pub fn insert(&mut self, index: String, facets: Option<ComputedFacets>) {
+        if let Some(facets) = facets {
+            self.0.insert(index, facets);
+        }
+    }
+
+    pub fn is_empty(&self) -> bool {
+        self.0.is_empty()
+    }
+
+    pub fn merge(
+        self,
+        MergeFacets { max_values_per_facet }: MergeFacets,
+        facet_order: BTreeMap<String, (String, OrderBy)>,
+    ) -> Option<ComputedFacets> {
+        if self.is_empty() {
+            return None;
+        }
+
+        let mut distribution: BTreeMap<String, _> = Default::default();
+        let mut stats: BTreeMap<String, FacetStats> = Default::default();
+
+        for facets_by_index in self.0.into_values() {
+            for (facet, index_distribution) in facets_by_index.distribution {
+                match distribution.entry(facet) {
+                    Entry::Vacant(entry) => {
+                        entry.insert(index_distribution);
+                    }
+                    Entry::Occupied(mut entry) => {
+                        let distribution = entry.get_mut();
+
+                        for (value, index_count) in index_distribution {
+                            distribution
+                                .entry(value)
+                                .and_modify(|count| *count += index_count)
+                                .or_insert(index_count);
+                        }
+                    }
+                }
+            }
+
+            for (facet, index_stats) in facets_by_index.stats {
+                match stats.entry(facet) {
+                    Entry::Vacant(entry) => {
+                        entry.insert(index_stats);
+                    }
+                    Entry::Occupied(mut entry) => {
+                        let stats = entry.get_mut();
+
+                        stats.min = f64::min(stats.min, index_stats.min);
+                        stats.max = f64::max(stats.max, index_stats.max);
+                    }
+                }
+            }
+        }
+
+        // fixup order
+        for (facet, values) in &mut distribution {
+            let order_by = facet_order.get(facet).map(|(_, order)| *order).unwrap_or_default();
+
+            match order_by {
+                OrderBy::Lexicographic => {
+                    values.sort_unstable_by(|left, _, right, _| left.cmp(right))
+                }
+                OrderBy::Count => {
+                    values.sort_unstable_by(|_, left, _, right| {
+                        left.cmp(right)
+                            // biggest first
+                            .reverse()
+                    })
+                }
+            }
+
+            if let Some(max_values_per_facet) = max_values_per_facet {
+                values.truncate(max_values_per_facet)
+            };
+        }
+
+        Some(ComputedFacets { distribution, stats })
+    }
+
+    pub(crate) fn append(&mut self, FederatedFacets(remote_facets_by_index): FederatedFacets) {
+        for (index, remote_facets) in remote_facets_by_index {
+            let merged_facets = self.0.entry(index).or_default();
+
+            for (remote_facet, remote_stats) in remote_facets.stats {
+                match merged_facets.stats.entry(remote_facet) {
+                    Entry::Vacant(vacant_entry) => {
+                        vacant_entry.insert(remote_stats);
+                    }
+                    Entry::Occupied(mut occupied_entry) => {
+                        let stats = occupied_entry.get_mut();
+                        stats.min = f64::min(stats.min, remote_stats.min);
+                        stats.max = f64::max(stats.max, remote_stats.max);
+                    }
+                }
+            }
+
+            for (remote_facet, remote_values) in remote_facets.distribution {
+                let merged_facet = merged_facets.distribution.entry(remote_facet).or_default();
+                for (remote_value, remote_count) in remote_values {
+                    let count = merged_facet.entry(remote_value).or_default();
+                    *count += remote_count;
+                }
+            }
+        }
+    }
+
+    pub fn sort_and_truncate(&mut self, facet_order: BTreeMap<String, (OrderByMap, usize)>) {
+        for (index, facets) in &mut self.0 {
+            let Some((order_by, max_values_per_facet)) = facet_order.get(index) else {
+                continue;
+            };
+            for (facet, values) in &mut facets.distribution {
+                match order_by.get(facet) {
+                    OrderBy::Lexicographic => {
+                        values.sort_unstable_by(|left, _, right, _| left.cmp(right))
+                    }
+                    OrderBy::Count => {
+                        values.sort_unstable_by(|_, left, _, right| {
+                            left.cmp(right)
+                                // biggest first
+                                .reverse()
+                        })
+                    }
+                }
+                values.truncate(*max_values_per_facet);
+            }
+        }
+    }
+}

crates/meilisearch/src/search/federated/weighted_scores.rs

@@ -0,0 +1,88 @@
+use std::cmp::Ordering;
+
+use meilisearch_types::milli::score_details::{self, WeightedScoreValue};
+
+pub fn compare(
+    mut left_it: impl Iterator<Item = WeightedScoreValue>,
+    left_weighted_global_score: f64,
+    mut right_it: impl Iterator<Item = WeightedScoreValue>,
+    right_weighted_global_score: f64,
+) -> Ordering {
+    loop {
+        let left = left_it.next();
+        let right = right_it.next();
+
+        match (left, right) {
+            (None, None) => return Ordering::Equal,
+            (None, Some(_)) => return Ordering::Less,
+            (Some(_), None) => return Ordering::Greater,
+            (
+                Some(
+                    WeightedScoreValue::WeightedScore(left) | WeightedScoreValue::VectorSort(left),
+                ),
+                Some(
+                    WeightedScoreValue::WeightedScore(right)
+                    | WeightedScoreValue::VectorSort(right),
+                ),
+            ) => {
+                if (left - right).abs() <= f64::EPSILON {
+                    continue;
+                }
+                return left.partial_cmp(&right).unwrap();
+            }
+            (
+                Some(WeightedScoreValue::Sort { asc: left_asc, value: left }),
+                Some(WeightedScoreValue::Sort { asc: right_asc, value: right }),
+            ) => {
+                if left_asc != right_asc {
+                    return left_weighted_global_score
+                        .partial_cmp(&right_weighted_global_score)
+                        .unwrap();
+                }
+                match score_details::compare_sort_values(left_asc, &left, &right) {
+                    Ordering::Equal => continue,
+                    order => return order,
+                }
+            }
+            (
+                Some(WeightedScoreValue::GeoSort { asc: left_asc, distance: left }),
+                Some(WeightedScoreValue::GeoSort { asc: right_asc, distance: right }),
+            ) => {
+                if left_asc != right_asc {
+                    continue;
+                }
+                match (left, right) {
+                    (None, None) => continue,
+                    (None, Some(_)) => return Ordering::Less,
+                    (Some(_), None) => return Ordering::Greater,
+                    (Some(left), Some(right)) => {
+                        if (left - right).abs() <= f64::EPSILON {
+                            continue;
+                        }
+                        return left.partial_cmp(&right).unwrap();
+                    }
+                }
+            }
+            // not comparable details, use global
+            (Some(WeightedScoreValue::WeightedScore(_)), Some(_))
+            | (Some(_), Some(WeightedScoreValue::WeightedScore(_)))
+            | (Some(WeightedScoreValue::VectorSort(_)), Some(_))
+            | (Some(_), Some(WeightedScoreValue::VectorSort(_)))
+            | (Some(WeightedScoreValue::GeoSort { .. }), Some(WeightedScoreValue::Sort { .. }))
+            | (Some(WeightedScoreValue::Sort { .. }), Some(WeightedScoreValue::GeoSort { .. })) => {
+                let left_count = left_it.count();
+                let right_count = right_it.count();
+                // compare how many remaining groups of rules each side has.
+                // the group with the most remaining groups wins.
+                return left_count
+                    .cmp(&right_count)
+                    // breaks ties with the global ranking score
+                    .then_with(|| {
+                        left_weighted_global_score
+                            .partial_cmp(&right_weighted_global_score)
+                            .unwrap()
+                    });
+            }
+        }
+    }
+}

crates/meilisearch/src/search/mod.rs

@@ -20,7 +20,7 @@ use meilisearch_types::milli::score_details::{ScoreDetails, ScoringStrategy};
 use meilisearch_types::milli::vector::parsed_vectors::ExplicitVectors;
 use meilisearch_types::milli::vector::Embedder;
 use meilisearch_types::milli::{
-    FacetValueHit, InternalError, OrderBy, SearchForFacetValues, TimeBudget,
+    FacetValueHit, InternalError, OrderBy, PatternMatch, SearchForFacetValues, TimeBudget,
 };
 use meilisearch_types::settings::DEFAULT_PAGINATION_MAX_TOTAL_HITS;
 use meilisearch_types::{milli, Document};
@@ -30,7 +30,7 @@ use milli::{
     MatchBounds, MatcherBuilder, SortError, TermsMatchingStrategy, DEFAULT_VALUES_PER_FACET,
 };
 use regex::Regex;
-use serde::Serialize;
+use serde::{Deserialize, Serialize};
 use serde_json::{json, Value};
 #[cfg(test)]
 mod mod_test;
@@ -41,7 +41,7 @@ use crate::error::MeilisearchHttpError;
 mod federated;
 pub use federated::{
     perform_federated_search, FederatedSearch, FederatedSearchResult, Federation,
-    FederationOptions, MergeFacets,
+    FederationOptions, MergeFacets, PROXY_SEARCH_HEADER, PROXY_SEARCH_HEADER_VALUE,
 };
 
 mod ranking_rules;
@@ -119,7 +119,7 @@ pub struct SearchQuery {
     pub locales: Option<Vec<Locale>>,
 }
 
-#[derive(Debug, Clone, Copy, PartialEq, Deserr, ToSchema)]
+#[derive(Debug, Clone, Copy, PartialEq, Deserr, ToSchema, Serialize)]
 #[deserr(try_from(f64) = TryFrom::try_from -> InvalidSearchRankingScoreThreshold)]
 pub struct RankingScoreThreshold(f64);
 impl std::convert::TryFrom<f64> for RankingScoreThreshold {
@@ -275,11 +275,13 @@ impl fmt::Debug for SearchQuery {
     }
 }
 
-#[derive(Debug, Clone, Default, PartialEq, Deserr, ToSchema)]
+#[derive(Debug, Clone, Default, PartialEq, Deserr, ToSchema, Serialize)]
 #[deserr(error = DeserrJsonError<InvalidSearchHybridQuery>, rename_all = camelCase, deny_unknown_fields)]
+#[serde(rename_all = "camelCase")]
 pub struct HybridQuery {
     #[deserr(default, error = DeserrJsonError<InvalidSearchSemanticRatio>, default)]
     #[schema(value_type = f32, default)]
+    #[serde(default)]
     pub semantic_ratio: SemanticRatio,
     #[deserr(error = DeserrJsonError<InvalidSearchEmbedder>)]
     pub embedder: String,
@@ -369,7 +371,7 @@ impl SearchKind {
     }
 }
 
-#[derive(Debug, Clone, Copy, PartialEq, Deserr)]
+#[derive(Debug, Clone, Copy, PartialEq, Deserr, Serialize)]
 #[deserr(try_from(f32) = TryFrom::try_from -> InvalidSearchSemanticRatio)]
 pub struct SemanticRatio(f32);
 
@@ -411,8 +413,9 @@ impl SearchQuery {
 // This struct contains the fields of `SearchQuery` inline.
 // This is because neither deserr nor serde support `flatten` when using `deny_unknown_fields.
 // The `From<SearchQueryWithIndex>` implementation ensures both structs remain up to date.
-#[derive(Debug, Clone, PartialEq, Deserr, ToSchema)]
+#[derive(Debug, Clone, Serialize, PartialEq, Deserr, ToSchema)]
 #[deserr(error = DeserrJsonError, rename_all = camelCase, deny_unknown_fields)]
+#[serde(rename_all = "camelCase")]
 #[schema(rename_all = "camelCase")]
 pub struct SearchQueryWithIndex {
     #[deserr(error = DeserrJsonError<InvalidIndexUid>, missing_field_error = DeserrJsonError::missing_index_uid)]
@@ -493,6 +496,72 @@ impl SearchQueryWithIndex {
         self.facets.as_deref().filter(|v| !v.is_empty())
     }
 
+    pub fn from_index_query_federation(
+        index_uid: IndexUid,
+        query: SearchQuery,
+        federation_options: Option<FederationOptions>,
+    ) -> Self {
+        let SearchQuery {
+            q,
+            vector,
+            hybrid,
+            offset,
+            limit,
+            page,
+            hits_per_page,
+            attributes_to_retrieve,
+            retrieve_vectors,
+            attributes_to_crop,
+            crop_length,
+            attributes_to_highlight,
+            show_matches_position,
+            show_ranking_score,
+            show_ranking_score_details,
+            filter,
+            sort,
+            distinct,
+            facets,
+            highlight_pre_tag,
+            highlight_post_tag,
+            crop_marker,
+            matching_strategy,
+            attributes_to_search_on,
+            ranking_score_threshold,
+            locales,
+        } = query;
+
+        SearchQueryWithIndex {
+            index_uid,
+            q,
+            vector,
+            hybrid,
+            offset: if offset == DEFAULT_SEARCH_OFFSET() { None } else { Some(offset) },
+            limit: if limit == DEFAULT_SEARCH_LIMIT() { None } else { Some(limit) },
+            page,
+            hits_per_page,
+            attributes_to_retrieve,
+            retrieve_vectors,
+            attributes_to_crop,
+            crop_length,
+            attributes_to_highlight,
+            show_ranking_score,
+            show_ranking_score_details,
+            show_matches_position,
+            filter,
+            sort,
+            distinct,
+            facets,
+            highlight_pre_tag,
+            highlight_post_tag,
+            crop_marker,
+            matching_strategy,
+            attributes_to_search_on,
+            ranking_score_threshold,
+            locales,
+            federation_options,
+        }
+    }
+
     pub fn into_index_query_federation(self) -> (IndexUid, SearchQuery, Option<FederationOptions>) {
         let SearchQueryWithIndex {
             index_uid,
@@ -620,8 +689,9 @@ impl TryFrom<Value> for ExternalDocumentId {
     }
 }
 
-#[derive(Debug, Copy, Clone, PartialEq, Eq, Deserr, ToSchema)]
+#[derive(Debug, Copy, Clone, PartialEq, Eq, Deserr, ToSchema, Serialize)]
 #[deserr(rename_all = camelCase)]
+#[serde(rename_all = "camelCase")]
 pub enum MatchingStrategy {
     /// Remove query words from last to first
     Last,
@@ -667,19 +737,19 @@ impl From<FacetValuesSort> for OrderBy {
     }
 }
 
-#[derive(Debug, Clone, Serialize, PartialEq, ToSchema)]
+#[derive(Debug, Clone, Serialize, Deserialize, PartialEq, ToSchema)]
 pub struct SearchHit {
     #[serde(flatten)]
     #[schema(additional_properties, inline, value_type = HashMap<String, Value>)]
     pub document: Document,
-    #[serde(rename = "_formatted", skip_serializing_if = "Document::is_empty")]
+    #[serde(default, rename = "_formatted", skip_serializing_if = "Document::is_empty")]
     #[schema(additional_properties, value_type = HashMap<String, Value>)]
     pub formatted: Document,
-    #[serde(rename = "_matchesPosition", skip_serializing_if = "Option::is_none")]
+    #[serde(default, rename = "_matchesPosition", skip_serializing_if = "Option::is_none")]
     pub matches_position: Option<MatchesPosition>,
-    #[serde(rename = "_rankingScore", skip_serializing_if = "Option::is_none")]
+    #[serde(default, rename = "_rankingScore", skip_serializing_if = "Option::is_none")]
     pub ranking_score: Option<f64>,
-    #[serde(rename = "_rankingScoreDetails", skip_serializing_if = "Option::is_none")]
+    #[serde(default, rename = "_rankingScoreDetails", skip_serializing_if = "Option::is_none")]
     pub ranking_score_details: Option<serde_json::Map<String, serde_json::Value>>,
 }
 
@@ -767,7 +837,7 @@ pub struct SearchResultWithIndex {
     pub result: SearchResult,
 }
 
-#[derive(Serialize, Debug, Clone, PartialEq, Eq, ToSchema)]
+#[derive(Serialize, Deserialize, Debug, Clone, PartialEq, Eq, ToSchema)]
 #[serde(untagged)]
 pub enum HitsInfo {
     #[serde(rename_all = "camelCase")]
@@ -778,7 +848,7 @@ pub enum HitsInfo {
     OffsetLimit { limit: usize, offset: usize, estimated_total_hits: usize },
 }
 
-#[derive(Serialize, Debug, Clone, PartialEq, ToSchema)]
+#[derive(Serialize, Deserialize, Debug, Clone, PartialEq, ToSchema)]
 pub struct FacetStats {
     pub min: f64,
     pub max: f64,
@@ -846,7 +916,7 @@ fn prepare_search<'t>(
                     let deadline = std::time::Instant::now() + std::time::Duration::from_secs(10);
 
                     embedder
-                        .embed_one(query.q.clone().unwrap(), Some(deadline))
+                        .embed_search(query.q.clone().unwrap(), Some(deadline))
                         .map_err(milli::vector::Error::from)
                         .map_err(milli::Error::from)?
                 }
@@ -1061,7 +1131,7 @@ pub fn perform_search(
     Ok(result)
 }
 
-#[derive(Debug, Clone, Default, Serialize, ToSchema)]
+#[derive(Debug, Clone, Default, Serialize, Deserialize, ToSchema)]
 pub struct ComputedFacets {
     #[schema(value_type = BTreeMap<String, BTreeMap<String, u64>>)]
     pub distribution: BTreeMap<String, IndexMap<String, u64>>,
@@ -1468,8 +1538,9 @@ pub fn perform_facet_search(
     // If the facet string is not localized, we **ignore** the locales provided by the user because the facet data has no locale.
     // If the user does not provide locales, we use the locales of the facet string.
     let localized_attributes = index.localized_attributes_rules(&rtxn)?.unwrap_or_default();
-    let localized_attributes_locales =
-        localized_attributes.into_iter().find(|attr| attr.match_str(&facet_name));
+    let localized_attributes_locales = localized_attributes
+        .into_iter()
+        .find(|attr| attr.match_str(&facet_name) == PatternMatch::Match);
     let locales = localized_attributes_locales.map(|attr| {
         attr.locales
             .into_iter()
@@ -1815,7 +1886,7 @@ fn format_fields(
             let locales = locales.or_else(|| {
                 localized_attributes
                     .iter()
-                    .find(|rule| rule.match_str(key))
+                    .find(|rule| rule.match_str(key) == PatternMatch::Match)
                     .map(LocalizedAttributesRule::locales)
             });
 

crates/meilisearch/tests/auth/api_keys.rs

@@ -421,7 +421,7 @@ async fn error_add_api_key_invalid_parameters_actions() {
     meili_snap::snapshot!(code, @"400 Bad Request");
     meili_snap::snapshot!(meili_snap::json_string!(response, { ".createdAt" => "[ignored]", ".updatedAt" => "[ignored]" }), @r###"
     {
-      "message": "Unknown value `doc.add` at `.actions[0]`: expected one of `*`, `search`, `documents.*`, `documents.add`, `documents.get`, `documents.delete`, `indexes.*`, `indexes.create`, `indexes.get`, `indexes.update`, `indexes.delete`, `indexes.swap`, `tasks.*`, `tasks.cancel`, `tasks.delete`, `tasks.get`, `settings.*`, `settings.get`, `settings.update`, `stats.*`, `stats.get`, `metrics.*`, `metrics.get`, `dumps.*`, `dumps.create`, `snapshots.*`, `snapshots.create`, `version`, `keys.create`, `keys.get`, `keys.update`, `keys.delete`, `experimental.get`, `experimental.update`",
+      "message": "Unknown value `doc.add` at `.actions[0]`: expected one of `*`, `search`, `documents.*`, `documents.add`, `documents.get`, `documents.delete`, `indexes.*`, `indexes.create`, `indexes.get`, `indexes.update`, `indexes.delete`, `indexes.swap`, `tasks.*`, `tasks.cancel`, `tasks.delete`, `tasks.get`, `settings.*`, `settings.get`, `settings.update`, `stats.*`, `stats.get`, `metrics.*`, `metrics.get`, `dumps.*`, `dumps.create`, `snapshots.*`, `snapshots.create`, `version`, `keys.create`, `keys.get`, `keys.update`, `keys.delete`, `experimental.get`, `experimental.update`, `network.get`, `network.update`",
       "code": "invalid_api_key_actions",
       "type": "invalid_request",
       "link": "https://docs.meilisearch.com/errors#invalid_api_key_actions"