Expose the three last batches timings

one

.github/release-draft-template.yml

@@ -1,29 +0,0 @@
-name-template: 'v$RESOLVED_VERSION'
-tag-template: 'v$RESOLVED_VERSION'
-exclude-labels:
-  - 'skip changelog'
-version-resolver:
-  minor:
-    labels:
-      - 'enhancement'
-  default: patch
-categories:
-  - title: '⚠️ Breaking changes'
-    label: 'breaking-change'
-  - title: '🚀 Enhancements'
-    label: 'enhancement'
-  - title: '🐛 Bug Fixes'
-    label: 'bug'
-  - title: '🔒 Security'
-    label: 'security'
-  - title: '⚙️ Maintenance/misc'
-    label:
-      - 'dependencies'
-      - 'maintenance'
-      - 'documentation'
-template: |
-  $CHANGES
-
-  ❤️  Huge thanks to our contributors: $CONTRIBUTORS.
-no-changes-template: 'Changes are coming soon 😎'
-sort-direction: 'ascending'

.github/workflows/release-drafter.yml

@@ -1,20 +0,0 @@
-name: Release Drafter
-
-permissions:
-  contents: read
-  pull-requests: write
-
-on:
-  push:
-    branches:
-      - main
-
-jobs:
-  update_release_draft:
-    runs-on: ubuntu-latest
-    steps:
-      - uses: release-drafter/release-drafter@v6
-        with:
-          config-name: release-draft-template.yml
-        env:
-          GITHUB_TOKEN: ${{ secrets.RELEASE_DRAFTER_TOKEN }}

Cargo.lock

@@ -453,9 +453,9 @@ checksum = "7c02d123df017efcdfbd739ef81735b36c5ba83ec3c59c80a9d7ecc718f92e50"
 
 [[package]]
 name = "arroy"
-version = "0.6.3"
+version = "0.6.4-nested-rtxns"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "8578a72223dfa13dfd9fc144d15260d134361789ebdea9b16e85a511edc73c7d"
+checksum = "fb8b6b34d9d83e3b837cb7f72a439dbd2293b102393c084af5e5b097212e1532"
 dependencies = [
  "bytemuck",
  "byteorder",
@@ -589,7 +589,7 @@ source = "git+https://github.com/meilisearch/bbqueue#cbb87cc707b5af415ef203bdaf2
 
 [[package]]
 name = "benchmarks"
-version = "1.22.1"
+version = "1.24.0"
 dependencies = [
  "anyhow",
  "bumpalo",
@@ -799,7 +799,7 @@ dependencies = [
 
 [[package]]
 name = "build-info"
-version = "1.22.1"
+version = "1.24.0"
 dependencies = [
  "anyhow",
  "time",
@@ -1075,9 +1075,9 @@ dependencies = [
 
 [[package]]
 name = "cellulite"
-version = "0.3.0"
+version = "0.3.1-nested-rtxns-2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "71a41aa2cd021bc3f23d97cc1e645848ca8c279fc757d1570ba7fe7ddc021290"
+checksum = "f22d721963ead1a144f10cb8b53dc9469e760723b069123c7c7bc675c7354270"
 dependencies = [
  "crossbeam",
  "geo",
@@ -1829,7 +1829,7 @@ dependencies = [
 
 [[package]]
 name = "dump"
-version = "1.22.1"
+version = "1.24.0"
 dependencies = [
  "anyhow",
  "big_s",
@@ -2072,7 +2072,7 @@ checksum = "37909eebbb50d72f9059c3b6d82c0463f2ff062c9e95845c43a6c9c0355411be"
 
 [[package]]
 name = "file-store"
-version = "1.22.1"
+version = "1.24.0"
 dependencies = [
  "tempfile",
  "thiserror 2.0.16",
@@ -2094,7 +2094,7 @@ dependencies = [
 
 [[package]]
 name = "filter-parser"
-version = "1.22.1"
+version = "1.24.0"
 dependencies = [
  "insta",
  "levenshtein_automata",
@@ -2122,7 +2122,7 @@ dependencies = [
 
 [[package]]
 name = "flatten-serde-json"
-version = "1.22.1"
+version = "1.24.0"
 dependencies = [
  "criterion",
  "serde_json",
@@ -2279,7 +2279,7 @@ dependencies = [
 
 [[package]]
 name = "fuzzers"
-version = "1.22.1"
+version = "1.24.0"
 dependencies = [
  "arbitrary",
  "bumpalo",
@@ -2758,9 +2758,9 @@ dependencies = [
 
 [[package]]
 name = "hannoy"
-version = "0.0.8"
+version = "0.0.9-nested-rtxns-2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "0dba13a271c49a119a97862ebf0a74131d879832868400d9fcd937b790058fdd"
+checksum = "06eda090938d9dcd568c8c2a5de383047ed9191578ebf4a342d2975d16e621f2"
 dependencies = [
  "bytemuck",
  "byteorder",
@@ -2838,9 +2838,9 @@ checksum = "2304e00983f87ffb38b55b444b5e3b60a884b5d30c0fca7d82fe33449bbe55ea"
 
 [[package]]
 name = "heed"
-version = "0.22.0"
+version = "0.22.1-nested-rtxns"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "6a56c94661ddfb51aa9cdfbf102cfcc340aa69267f95ebccc4af08d7c530d393"
+checksum = "0ff115ba5712b1f1fc7617b195f5c2f139e29c397ff79da040cd19db75ccc240"
 dependencies = [
  "bitflags 2.9.4",
  "byteorder",
@@ -2850,7 +2850,6 @@ dependencies = [
  "lmdb-master-sys",
  "once_cell",
  "page_size",
- "serde",
  "synchronoise",
  "url",
 ]
@@ -2968,6 +2967,12 @@ version = "1.0.3"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "df3b46402a9d5adb4c86a0cf463f42e19994e3ee891101b1841f30a545cb49a9"
 
+[[package]]
+name = "humantime"
+version = "2.3.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "135b12329e5e3ce057a9f972339ea52bc954fe1e9358ef27f95e89716fbc5424"
+
 [[package]]
 name = "hyper"
 version = "1.7.0"
@@ -3234,7 +3239,7 @@ dependencies = [
 
 [[package]]
 name = "index-scheduler"
-version = "1.22.1"
+version = "1.24.0"
 dependencies = [
  "anyhow",
  "backoff",
@@ -3488,7 +3493,7 @@ dependencies = [
 
 [[package]]
 name = "json-depth-checker"
-version = "1.22.1"
+version = "1.24.0"
 dependencies = [
  "criterion",
  "serde_json",
@@ -3889,9 +3894,9 @@ checksum = "241eaef5fd12c88705a01fc1066c48c4b36e0dd4377dcdc7ec3942cea7a69956"
 
 [[package]]
 name = "lmdb-master-sys"
-version = "0.2.5"
+version = "0.2.6-nested-rtxns"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "864808e0b19fb6dd3b70ba94ee671b82fce17554cf80aeb0a155c65bb08027df"
+checksum = "f4ff85130e3c994b36877045fbbb138d521dea7197bfc19dc3d5d95101a8e20a"
 dependencies = [
  "cc",
  "doxygen-rs",
@@ -3997,7 +4002,7 @@ checksum = "490cc448043f947bae3cbee9c203358d62dbee0db12107a74be5c30ccfd09771"
 
 [[package]]
 name = "meili-snap"
-version = "1.22.1"
+version = "1.24.0"
 dependencies = [
  "insta",
  "md5",
@@ -4008,7 +4013,7 @@ dependencies = [
 
 [[package]]
 name = "meilisearch"
-version = "1.22.1"
+version = "1.24.0"
 dependencies = [
  "actix-cors",
  "actix-http",
@@ -4037,6 +4042,7 @@ dependencies = [
  "futures",
  "futures-util",
  "hex",
+ "humantime",
  "index-scheduler",
  "indexmap",
  "insta",
@@ -4105,7 +4111,7 @@ dependencies = [
 
 [[package]]
 name = "meilisearch-auth"
-version = "1.22.1"
+version = "1.24.0"
 dependencies = [
  "base64 0.22.1",
  "enum-iterator",
@@ -4124,7 +4130,7 @@ dependencies = [
 
 [[package]]
 name = "meilisearch-types"
-version = "1.22.1"
+version = "1.24.0"
 dependencies = [
  "actix-web",
  "anyhow",
@@ -4159,7 +4165,7 @@ dependencies = [
 
 [[package]]
 name = "meilitool"
-version = "1.22.1"
+version = "1.24.0"
 dependencies = [
  "anyhow",
  "clap",
@@ -4193,7 +4199,7 @@ dependencies = [
 
 [[package]]
 name = "milli"
-version = "1.22.1"
+version = "1.24.0"
 dependencies = [
  "allocator-api2 0.3.1",
  "arroy",
@@ -4774,7 +4780,7 @@ checksum = "9b4f627cb1b25917193a259e49bdad08f671f8d9708acfd5fe0a8c1455d87220"
 
 [[package]]
 name = "permissive-json-pointer"
-version = "1.22.1"
+version = "1.24.0"
 dependencies = [
  "big_s",
  "serde_json",
@@ -7821,7 +7827,7 @@ dependencies = [
 
 [[package]]
 name = "xtask"
-version = "1.22.1"
+version = "1.24.0"
 dependencies = [
  "anyhow",
  "build-info",

Cargo.toml

@@ -23,7 +23,7 @@ members = [
 ]
 
 [workspace.package]
-version = "1.22.1"
+version = "1.24.0"
 authors = [
     "Quentin de Quelen <quentin@dequelen.me>",
     "Clément Renault <clement@meilisearch.com>",

Dockerfile

@@ -1,5 +1,5 @@
 # Compile
-FROM    rust:1.89-alpine3.20 AS compiler
+FROM    rust:1.89-alpine3.22 AS compiler
 
 RUN     apk add -q --no-cache build-base openssl-dev
 
@@ -20,7 +20,7 @@ RUN     set -eux; \
         cargo build --release -p meilisearch -p meilitool
 
 # Run
-FROM    alpine:3.20
+FROM    alpine:3.22
 LABEL   org.opencontainers.image.source="https://github.com/meilisearch/meilisearch"
 
 ENV     MEILI_HTTP_ADDR 0.0.0.0:7700

LICENSE

@@ -1,29 +1,9 @@
-MIT License
+# License
 
 Copyright (c) 2019-2025 Meili SAS
 
-Permission is hereby granted, free of charge, to any person obtaining a copy
-of this software and associated documentation files (the "Software"), to deal
-in the Software without restriction, including without limitation the rights
-to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
-copies of the Software, and to permit persons to whom the Software is
-furnished to do so, subject to the following conditions:
+Part of this work fall under the Meilisearch Enterprise Edition (EE) and are licensed under the Business Source License 1.1, please refer to [LICENSE-EE](./LICENSE-EE) for details.
 
-The above copyright notice and this permission notice shall be included in all
-copies or substantial portions of the Software.
+The other parts of this work are licensed under the [MIT license](./LICENSE-MIT).
 
-THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
-OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
-SOFTWARE.
-
----
-
-🔒 Meilisearch Enterprise Edition (EE)
-
-Certain parts of this codebase are not licensed under the MIT license and governed by the Business Source License 1.1.
-
-See the LICENSE-EE file for details.
+`SPDX-License-Identifier: MIT AND BUSL-1.1`

LICENSE-MIT

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2019-2025 Meili SAS
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

crates/dump/src/lib.rs

@@ -158,6 +158,9 @@ pub enum KindDump {
     UpgradeDatabase {
         from: (u32, u32, u32),
     },
+    IndexCompaction {
+        index_uid: String,
+    },
 }
 
 impl From<Task> for TaskDump {
@@ -240,6 +243,9 @@ impl From<KindWithContent> for KindDump {
             KindWithContent::UpgradeDatabase { from: version } => {
                 KindDump::UpgradeDatabase { from: version }
             }
+            KindWithContent::IndexCompaction { index_uid } => {
+                KindDump::IndexCompaction { index_uid }
+            }
         }
     }
 }

crates/file-store/src/lib.rs

@@ -33,10 +33,6 @@ impl FileStore {
         std::fs::create_dir_all(&path)?;
         Ok(FileStore { path })
     }
-
-    pub fn path(&self) -> &Path {
-        &self.path
-    }
 }
 
 impl FileStore {

crates/index-scheduler/src/dump.rs

@@ -234,6 +234,9 @@ impl<'a> Dump<'a> {
                     }
                 }
                 KindDump::UpgradeDatabase { from } => KindWithContent::UpgradeDatabase { from },
+                KindDump::IndexCompaction { index_uid } => {
+                    KindWithContent::IndexCompaction { index_uid }
+                }
             },
         };
 

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

@@ -1,4 +1,4 @@
-use std::path::{Path, PathBuf};
+use std::path::PathBuf;
 use std::sync::{Arc, RwLock};
 use std::time::Duration;
 use std::{fs, thread};
@@ -199,7 +199,7 @@ impl IndexMapper {
                 let uuid = Uuid::new_v4();
                 self.index_mapping.put(&mut wtxn, name, &uuid)?;
 
-                let index_path = self.base_path.join(uuid.to_string());
+                let index_path = self.index_path(uuid);
                 fs::create_dir_all(&index_path)?;
 
                 // Error if the UUIDv4 somehow already exists in the map, since it should be fresh.
@@ -286,7 +286,7 @@ impl IndexMapper {
         };
 
         let index_map = self.index_map.clone();
-        let index_path = self.base_path.join(uuid.to_string());
+        let index_path = self.index_path(uuid);
         let index_name = name.to_string();
         thread::Builder::new()
             .name(String::from("index_deleter"))
@@ -341,6 +341,26 @@ impl IndexMapper {
         Ok(())
     }
 
+    /// Closes the specified index.
+    ///
+    /// This operation involves closing the underlying environment and so can take a long time to complete.
+    ///
+    /// # Panics
+    ///
+    /// - If the Index corresponding to the passed name is concurrently being deleted/resized or cannot be found in the
+    ///   in memory hash map.
+    pub fn close_index(&self, rtxn: &RoTxn, name: &str) -> Result<()> {
+        let uuid = self
+            .index_mapping
+            .get(rtxn, name)?
+            .ok_or_else(|| Error::IndexNotFound(name.to_string()))?;
+
+        // We remove the index from the in-memory index map.
+        self.index_map.write().unwrap().close_for_resize(&uuid, self.enable_mdb_writemap, 0);
+
+        Ok(())
+    }
+
     /// Return an index, may open it if it wasn't already opened.
     pub fn index(&self, rtxn: &RoTxn, name: &str) -> Result<Index> {
         if let Some((current_name, current_index)) =
@@ -388,7 +408,7 @@ impl IndexMapper {
                     } else {
                         continue;
                     };
-                    let index_path = self.base_path.join(uuid.to_string());
+                    let index_path = self.index_path(uuid);
                     // take the lock to reopen the environment.
                     reopen
                         .reopen(&mut self.index_map.write().unwrap(), &index_path)
@@ -405,7 +425,7 @@ impl IndexMapper {
                     // if it's not already there.
                     match index_map.get(&uuid) {
                         Missing => {
-                            let index_path = self.base_path.join(uuid.to_string());
+                            let index_path = self.index_path(uuid);
 
                             break index_map
                                 .create(
@@ -432,6 +452,14 @@ impl IndexMapper {
         Ok(index)
     }
 
+    /// Returns the path of the index.
+    ///
+    /// The folder located at this path is containing the data.mdb,
+    /// the lock.mdb and an optional data.mdb.cpy file.
+    pub fn index_path(&self, uuid: Uuid) -> PathBuf {
+        self.base_path.join(uuid.to_string())
+    }
+
     pub fn rollback_index(
         &self,
         rtxn: &RoTxn,
@@ -472,7 +500,7 @@ impl IndexMapper {
             };
         }
 
-        let index_path = self.base_path.join(uuid.to_string());
+        let index_path = self.index_path(uuid);
         Index::rollback(milli::heed::EnvOpenOptions::new().read_txn_without_tls(), index_path, to)
             .map_err(|err| crate::Error::from_milli(err, Some(name.to_string())))
     }
@@ -591,8 +619,4 @@ impl IndexMapper {
     pub fn set_currently_updating_index(&self, index: Option<(String, Index)>) {
         *self.currently_updating_index.write().unwrap() = index;
     }
-
-    pub fn base_path(&self) -> &Path {
-        &self.base_path
-    }
 }

crates/index-scheduler/src/insta_snapshot.rs

@@ -317,6 +317,9 @@ fn snapshot_details(d: &Details) -> String {
         Details::UpgradeDatabase { from, to } => {
             format!("{{ from: {from:?}, to: {to:?} }}")
         }
+        Details::IndexCompaction { index_uid, pre_compaction_size, post_compaction_size } => {
+            format!("{{ index_uid: {index_uid:?}, pre_compaction_size: {pre_compaction_size:?}, post_compaction_size: {post_compaction_size:?} }}")
+        }
     }
 }
 

crates/index-scheduler/src/processing.rs

@@ -75,6 +75,7 @@ make_enum_progress! {
     pub enum TaskCancelationProgress {
         RetrievingTasks,
         CancelingUpgrade,
+        CleaningCompactionLeftover,
         UpdatingTasks,
     }
 }
@@ -138,6 +139,17 @@ make_enum_progress! {
     }
 }
 
+make_enum_progress! {
+    pub enum IndexCompaction {
+        RetrieveTheIndex,
+        CreateTemporaryFile,
+        CopyAndCompactTheIndex,
+        PersistTheCompactedIndex,
+        CloseTheIndex,
+        ReopenTheIndex,
+    }
+}
+
 make_enum_progress! {
     pub enum InnerSwappingTwoIndexes {
         RetrieveTheTasks,

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

@@ -68,13 +68,14 @@ impl From<KindWithContent> for AutobatchKind {
             KindWithContent::IndexCreation { .. } => AutobatchKind::IndexCreation,
             KindWithContent::IndexUpdate { .. } => AutobatchKind::IndexUpdate,
             KindWithContent::IndexSwap { .. } => AutobatchKind::IndexSwap,
-            KindWithContent::TaskCancelation { .. }
+            KindWithContent::IndexCompaction { .. }
+            | KindWithContent::TaskCancelation { .. }
             | KindWithContent::TaskDeletion { .. }
             | KindWithContent::DumpCreation { .. }
             | KindWithContent::Export { .. }
             | KindWithContent::UpgradeDatabase { .. }
             | KindWithContent::SnapshotCreation => {
-                panic!("The autobatcher should never be called with tasks that don't apply to an index.")
+                panic!("The autobatcher should never be called with tasks with special priority or that don't apply to an index.")
             }
         }
     }
@@ -287,8 +288,10 @@ impl BatchKind {
         };
 
         match (self, autobatch_kind) {
-            // We don't batch any of these operations  
-            (this, K::IndexCreation | K::IndexUpdate | K::IndexSwap | K::DocumentEdition) => Break((this, BatchStopReason::TaskCannotBeBatched { kind, id })),
+            // We don't batch any of these operations
+            (this, K::IndexCreation | K::IndexUpdate | K::IndexSwap | K::DocumentEdition) => {
+                Break((this, BatchStopReason::TaskCannotBeBatched { kind, id }))
+            },
             // We must not batch tasks that don't have the same index creation rights if the index doesn't already exists.
             (this, kind) if !index_already_exists && this.allow_index_creation() == Some(false) && kind.allow_index_creation() == Some(true) => {
                 Break((this, BatchStopReason::IndexCreationMismatch { id }))

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

@@ -55,6 +55,10 @@ pub(crate) enum Batch {
     UpgradeDatabase {
         tasks: Vec<Task>,
     },
+    IndexCompaction {
+        index_uid: String,
+        task: Task,
+    },
 }
 
 #[derive(Debug)]
@@ -110,7 +114,8 @@ impl Batch {
             | Batch::Dump(task)
             | Batch::IndexCreation { task, .. }
             | Batch::Export { task }
-            | Batch::IndexUpdate { task, .. } => {
+            | Batch::IndexUpdate { task, .. }
+            | Batch::IndexCompaction { task, .. } => {
                 RoaringBitmap::from_sorted_iter(std::iter::once(task.uid)).unwrap()
             }
             Batch::SnapshotCreation(tasks)
@@ -155,7 +160,8 @@ impl Batch {
             IndexOperation { op, .. } => Some(op.index_uid()),
             IndexCreation { index_uid, .. }
             | IndexUpdate { index_uid, .. }
-            | IndexDeletion { index_uid, .. } => Some(index_uid),
+            | IndexDeletion { index_uid, .. }
+            | IndexCompaction { index_uid, .. } => Some(index_uid),
         }
     }
 }
@@ -175,6 +181,7 @@ impl fmt::Display for Batch {
             Batch::IndexUpdate { .. } => f.write_str("IndexUpdate")?,
             Batch::IndexDeletion { .. } => f.write_str("IndexDeletion")?,
             Batch::IndexSwap { .. } => f.write_str("IndexSwap")?,
+            Batch::IndexCompaction { .. } => f.write_str("IndexCompaction")?,
             Batch::Export { .. } => f.write_str("Export")?,
             Batch::UpgradeDatabase { .. } => f.write_str("UpgradeDatabase")?,
         };
@@ -512,17 +519,33 @@ impl IndexScheduler {
             return Ok(Some((Batch::TaskDeletions(tasks), current_batch)));
         }
 
-        // 3. we batch the export.
+        // 3. we get the next task to compact
+        let to_compact = self.queue.tasks.get_kind(rtxn, Kind::IndexCompaction)? & enqueued;
+        if let Some(task_id) = to_compact.min() {
+            let mut task =
+                self.queue.tasks.get_task(rtxn, task_id)?.ok_or(Error::CorruptedTaskQueue)?;
+            current_batch.processing(Some(&mut task));
+            current_batch.reason(BatchStopReason::TaskCannotBeBatched {
+                kind: Kind::IndexCompaction,
+                id: task_id,
+            });
+            let index_uid =
+                task.index_uid().expect("Compaction task must have an index uid").to_owned();
+            return Ok(Some((Batch::IndexCompaction { index_uid, task }, current_batch)));
+        }
+
+        // 4. we batch the export.
         let to_export = self.queue.tasks.get_kind(rtxn, Kind::Export)? & enqueued;
         if !to_export.is_empty() {
             let task_id = to_export.iter().next().expect("There must be at least one export task");
             let mut task = self.queue.tasks.get_task(rtxn, task_id)?.unwrap();
             current_batch.processing([&mut task]);
-            current_batch.reason(BatchStopReason::TaskKindCannotBeBatched { kind: Kind::Export });
+            current_batch
+                .reason(BatchStopReason::TaskCannotBeBatched { kind: Kind::Export, id: task_id });
             return Ok(Some((Batch::Export { task }, current_batch)));
         }
 
-        // 4. we batch the snapshot.
+        // 5. we batch the snapshot.
         let to_snapshot = self.queue.tasks.get_kind(rtxn, Kind::SnapshotCreation)? & enqueued;
         if !to_snapshot.is_empty() {
             let mut tasks = self.queue.tasks.get_existing_tasks(rtxn, to_snapshot)?;
@@ -532,7 +555,7 @@ impl IndexScheduler {
             return Ok(Some((Batch::SnapshotCreation(tasks), current_batch)));
         }
 
-        // 5. we batch the dumps.
+        // 6. we batch the dumps.
         let to_dump = self.queue.tasks.get_kind(rtxn, Kind::DumpCreation)? & enqueued;
         if let Some(to_dump) = to_dump.min() {
             let mut task =
@@ -545,7 +568,7 @@ impl IndexScheduler {
             return Ok(Some((Batch::Dump(task), current_batch)));
         }
 
-        // 6. We make a batch from the unprioritised tasks. Start by taking the next enqueued task.
+        // 7. We make a batch from the unprioritised tasks. Start by taking the next enqueued task.
         let task_id = if let Some(task_id) = enqueued.min() { task_id } else { return Ok(None) };
         let mut task =
             self.queue.tasks.get_task(rtxn, task_id)?.ok_or(Error::CorruptedTaskQueue)?;

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

@@ -50,11 +50,6 @@ impl MustStopProcessing {
     pub fn reset(&self) {
         self.0.store(false, Ordering::Relaxed);
     }
-
-    pub fn as_lambda(&self) -> impl Fn() -> bool + Send + Sync + 'static {
-        let clone = self.clone();
-        move || clone.get()
-    }
 }
 
 pub struct Scheduler {

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

@@ -1,22 +1,27 @@
 use std::collections::{BTreeSet, HashMap, HashSet};
+use std::fs::{remove_file, File};
+use std::io::{ErrorKind, Seek, SeekFrom};
 use std::panic::{catch_unwind, AssertUnwindSafe};
 use std::sync::atomic::Ordering;
 
+use byte_unit::Byte;
 use meilisearch_types::batches::{BatchEnqueuedAt, BatchId};
 use meilisearch_types::heed::{RoTxn, RwTxn};
+use meilisearch_types::milli::heed::CompactionOption;
 use meilisearch_types::milli::progress::{Progress, VariableNameStep};
 use meilisearch_types::milli::{self, ChannelCongestion};
 use meilisearch_types::tasks::{Details, IndexSwap, Kind, KindWithContent, Status, Task};
 use meilisearch_types::versioning::{VERSION_MAJOR, VERSION_MINOR, VERSION_PATCH};
 use milli::update::Settings as MilliSettings;
 use roaring::RoaringBitmap;
+use tempfile::{PersistError, TempPath};
 use time::OffsetDateTime;
 
 use super::create_batch::Batch;
 use crate::processing::{
     AtomicBatchStep, AtomicTaskStep, CreateIndexProgress, DeleteIndexProgress, FinalizingIndexStep,
-    InnerSwappingTwoIndexes, SwappingTheIndexes, TaskCancelationProgress, TaskDeletionProgress,
-    UpdateIndexProgress,
+    IndexCompaction, InnerSwappingTwoIndexes, SwappingTheIndexes, TaskCancelationProgress,
+    TaskDeletionProgress, UpdateIndexProgress,
 };
 use crate::utils::{
     self, remove_n_tasks_datetime_earlier_than, remove_task_datetime, swap_index_uid_in_task,
@@ -24,6 +29,9 @@ use crate::utils::{
 };
 use crate::{Error, IndexScheduler, Result, TaskId};
 
+/// The name of the copy of the data.mdb file used during compaction.
+const DATA_MDB_COPY_NAME: &str = "data.mdb.cpy";
+
 #[derive(Debug, Default)]
 pub struct ProcessBatchInfo {
     /// The write channel congestion. None when unavailable: settings update.
@@ -418,6 +426,47 @@ impl IndexScheduler {
                 task.status = Status::Succeeded;
                 Ok((vec![task], ProcessBatchInfo::default()))
             }
+            Batch::IndexCompaction { index_uid: _, mut task } => {
+                let KindWithContent::IndexCompaction { index_uid } = &task.kind else {
+                    unreachable!()
+                };
+
+                let rtxn = self.env.read_txn()?;
+                let ret = catch_unwind(AssertUnwindSafe(|| {
+                    self.apply_compaction(&rtxn, &progress, index_uid)
+                }));
+
+                let (pre_size, post_size) = match ret {
+                    Ok(Ok(stats)) => stats,
+                    Ok(Err(Error::AbortedTask)) => return Err(Error::AbortedTask),
+                    Ok(Err(e)) => return Err(e),
+                    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::Export(Box::new(Error::ProcessBatchPanicked(
+                            msg.to_string(),
+                        ))));
+                    }
+                };
+
+                task.status = Status::Succeeded;
+                if let Some(Details::IndexCompaction {
+                    index_uid: _,
+                    pre_compaction_size,
+                    post_compaction_size,
+                }) = task.details.as_mut()
+                {
+                    *pre_compaction_size = Some(Byte::from_u64(pre_size));
+                    *post_compaction_size = Some(Byte::from_u64(post_size));
+                }
+
+                Ok((vec![task], ProcessBatchInfo::default()))
+            }
             Batch::Export { mut task } => {
                 let KindWithContent::Export { url, api_key, payload_size, indexes } = &task.kind
                 else {
@@ -493,6 +542,92 @@ impl IndexScheduler {
         }
     }
 
+    fn apply_compaction(
+        &self,
+        rtxn: &RoTxn,
+        progress: &Progress,
+        index_uid: &str,
+    ) -> Result<(u64, u64)> {
+        // 1. Verify that the index exists
+        if !self.index_mapper.index_exists(rtxn, index_uid)? {
+            return Err(Error::IndexNotFound(index_uid.to_owned()));
+        }
+
+        // 2. We retrieve the index and create a temporary file in the index directory
+        progress.update_progress(IndexCompaction::RetrieveTheIndex);
+        let index = self.index_mapper.index(rtxn, index_uid)?;
+
+        // the index operation can take a long time, so save this handle to make it available to the search for the duration of the tick
+        self.index_mapper
+            .set_currently_updating_index(Some((index_uid.to_string(), index.clone())));
+
+        progress.update_progress(IndexCompaction::CreateTemporaryFile);
+        let src_path = index.path().join("data.mdb");
+        let pre_size = std::fs::metadata(&src_path)?.len();
+
+        let dst_path = TempPath::from_path(index.path().join(DATA_MDB_COPY_NAME));
+        let file = File::create(&dst_path)?;
+        let mut file = tempfile::NamedTempFile::from_parts(file, dst_path);
+
+        // 3. We copy the index data to the temporary file
+        progress.update_progress(IndexCompaction::CopyAndCompactTheIndex);
+        index
+            .copy_to_file(file.as_file_mut(), CompactionOption::Enabled)
+            .map_err(|error| Error::Milli { error, index_uid: Some(index_uid.to_string()) })?;
+        // ...and reset the file position as specified in the documentation
+        file.seek(SeekFrom::Start(0))?;
+
+        // 4. We replace the index data file with the temporary file
+        progress.update_progress(IndexCompaction::PersistTheCompactedIndex);
+        match file.persist(src_path) {
+            Ok(file) => file.sync_all()?,
+            // TODO see if we have a _resource busy_ error and probably handle this by:
+            //      1. closing the index, 2. replacing and 3. reopening it
+            Err(PersistError { error, file: _ }) => return Err(Error::IoError(error)),
+        };
+
+        // 5. Prepare to close the index
+        progress.update_progress(IndexCompaction::CloseTheIndex);
+
+        // unmark that the index is the processing one so we don't keep a handle to it, preventing its closing
+        self.index_mapper.set_currently_updating_index(None);
+
+        self.index_mapper.close_index(rtxn, index_uid)?;
+        drop(index);
+
+        progress.update_progress(IndexCompaction::ReopenTheIndex);
+        // 6. Reopen the index
+        // The index will use the compacted data file when being reopened
+        let index = self.index_mapper.index(rtxn, index_uid)?;
+
+        // if the update processed successfully, we're going to store the new
+        // stats of the index. Since the tasks have already been processed and
+        // this is a non-critical operation. If it fails, we should not fail
+        // the entire batch.
+        let res = || -> Result<_> {
+            let mut wtxn = self.env.write_txn()?;
+            let index_rtxn = index.read_txn()?;
+            let stats = crate::index_mapper::IndexStats::new(&index, &index_rtxn)
+                .map_err(|e| Error::from_milli(e, Some(index_uid.to_string())))?;
+            self.index_mapper.store_stats_of(&mut wtxn, index_uid, &stats)?;
+            wtxn.commit()?;
+            Ok(stats.database_size)
+        }();
+
+        let post_size = match res {
+            Ok(post_size) => post_size,
+            Err(e) => {
+                tracing::error!(
+                    error = &e as &dyn std::error::Error,
+                    "Could not write the stats of the index"
+                );
+                0
+            }
+        };
+
+        Ok((pre_size, post_size))
+    }
+
     /// Swap the index `lhs` with the index `rhs`.
     fn apply_index_swap(
         &self,
@@ -780,9 +915,10 @@ impl IndexScheduler {
 
         let enqueued_tasks = &self.queue.tasks.get_status(rtxn, Status::Enqueued)?;
 
-        // 0. Check if any upgrade task was matched.
+        // 0. Check if any upgrade or compaction tasks were matched.
         //    If so, we cancel all the failed or enqueued upgrade tasks.
         let upgrade_tasks = &self.queue.tasks.get_kind(rtxn, Kind::UpgradeDatabase)?;
+        let compaction_tasks = &self.queue.tasks.get_kind(rtxn, Kind::IndexCompaction)?;
         let is_canceling_upgrade = !matched_tasks.is_disjoint(upgrade_tasks);
         if is_canceling_upgrade {
             let failed_tasks = self.queue.tasks.get_status(rtxn, Status::Failed)?;
@@ -847,7 +983,33 @@ impl IndexScheduler {
             }
         }
 
-        // 3. We now have a list of tasks to cancel, cancel them
+        // 3. If we are cancelling a compaction task, remove the tempfiles after incomplete compactions
+        for compaction_task in &tasks_to_cancel & compaction_tasks {
+            progress.update_progress(TaskCancelationProgress::CleaningCompactionLeftover);
+            let task = self.queue.tasks.get_task(rtxn, compaction_task)?.unwrap();
+            let Some(Details::IndexCompaction {
+                index_uid,
+                pre_compaction_size: _,
+                post_compaction_size: _,
+            }) = task.details
+            else {
+                unreachable!("wrong details for compaction task {compaction_task}")
+            };
+
+            let index_path = match self.index_mapper.index_mapping.get(rtxn, &index_uid)? {
+                Some(index_uuid) => self.index_mapper.index_path(index_uuid),
+                None => continue,
+            };
+
+            if let Err(e) = remove_file(index_path.join(DATA_MDB_COPY_NAME)) {
+                match e.kind() {
+                    ErrorKind::NotFound => (),
+                    _ => return Err(Error::IoError(e)),
+                }
+            }
+        }
+
+        // 4. We now have a list of tasks to cancel, cancel them
         let (task_progress, progress_obj) = AtomicTaskStep::new(tasks_to_cancel.len() as u32);
         progress.update_progress(progress_obj);
 

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

@@ -4,7 +4,6 @@ use std::sync::atomic::Ordering;
 
 use meilisearch_types::heed::CompactionOption;
 use meilisearch_types::milli::progress::{Progress, VariableNameStep};
-use meilisearch_types::milli::InternalError;
 use meilisearch_types::tasks::{Status, Task};
 use meilisearch_types::{compression, VERSION_FILE_NAME};
 
@@ -77,22 +76,6 @@ unsafe fn remove_tasks(
 
 impl IndexScheduler {
     pub(super) fn process_snapshot(
-        &self,
-        progress: Progress,
-        tasks: Vec<Task>,
-    ) -> Result<Vec<Task>> {
-        let compaction_option = if self.scheduler.experimental_no_snapshot_compaction {
-            CompactionOption::Disabled
-        } else {
-            CompactionOption::Enabled
-        };
-        match compaction_option {
-            CompactionOption::Enabled => self.process_snapshot_with_temp(progress, tasks),
-            CompactionOption::Disabled => self.process_snapshot_with_pipe(progress, tasks),
-        }
-    }
-
-    fn process_snapshot_with_temp(
         &self,
         progress: Progress,
         mut tasks: Vec<Task>,
@@ -122,8 +105,12 @@ impl IndexScheduler {
         progress.update_progress(SnapshotCreationProgress::SnapshotTheIndexScheduler);
         let dst = temp_snapshot_dir.path().join("tasks");
         fs::create_dir_all(&dst)?;
-
-        self.env.copy_to_path(dst.join("data.mdb"), CompactionOption::Enabled)?;
+        let compaction_option = if self.scheduler.experimental_no_snapshot_compaction {
+            CompactionOption::Disabled
+        } else {
+            CompactionOption::Enabled
+        };
+        self.env.copy_to_path(dst.join("data.mdb"), compaction_option)?;
 
         // 2.2 Remove the current snapshot tasks
         //
@@ -174,7 +161,7 @@ impl IndexScheduler {
             let dst = temp_snapshot_dir.path().join("indexes").join(uuid.to_string());
             fs::create_dir_all(&dst)?;
             index
-                .copy_to_path(dst.join("data.mdb"), CompactionOption::Enabled)
+                .copy_to_path(dst.join("data.mdb"), compaction_option)
                 .map_err(|e| Error::from_milli(e, Some(name.to_string())))?;
         }
 
@@ -184,7 +171,7 @@ impl IndexScheduler {
         progress.update_progress(SnapshotCreationProgress::SnapshotTheApiKeys);
         let dst = temp_snapshot_dir.path().join("auth");
         fs::create_dir_all(&dst)?;
-        self.scheduler.auth_env.copy_to_path(dst.join("data.mdb"), CompactionOption::Enabled)?;
+        self.scheduler.auth_env.copy_to_path(dst.join("data.mdb"), compaction_option)?;
 
         // 5. Copy and tarball the flat snapshot
         progress.update_progress(SnapshotCreationProgress::CreateTheTarball);
@@ -219,138 +206,4 @@ impl IndexScheduler {
 
         Ok(tasks)
     }
-
-    fn process_snapshot_with_pipe(
-        &self,
-        progress: Progress,
-        mut tasks: Vec<Task>,
-    ) -> Result<Vec<Task>> {
-        progress.update_progress(SnapshotCreationProgress::StartTheSnapshotCreation);
-        let must_stop_processing = &self.scheduler.must_stop_processing;
-        let abort_no_index = Err(Error::from_milli(InternalError::AbortedIndexation.into(), None));
-
-        fs::create_dir_all(&self.scheduler.snapshots_path)?;
-
-        // 1. Find the base path and original name of the database
-
-        // TODO find a better way to get this path
-        let mut base_path = self.env.path().to_owned();
-        base_path.pop();
-        let base_path = base_path;
-        let db_name = base_path.file_name().and_then(OsStr::to_str).unwrap_or("data.ms");
-
-        // 2. Start the tarball builder. The tarball will be created on another thread from piped data.
-
-        let mut builder = compression::PipedArchiveBuilder::new(
-            self.scheduler.snapshots_path.clone(),
-            base_path.clone(),
-        )?;
-
-        // 3. Snapshot the VERSION file
-        builder.add_file_to_archive(self.scheduler.version_file_path.clone())?;
-        if must_stop_processing.get() {
-            return abort_no_index;
-        }
-
-        // 4. Snapshot the index-scheduler LMDB env
-        //
-        // When we call copy_to_path, LMDB opens a read transaction by itself,
-        // we can't provide our own. It is an issue as we would like to know
-        // the update files to copy but new ones can be enqueued between the copy
-        // of the env and the new transaction we open to retrieve the enqueued tasks.
-        // So we prefer opening a new transaction after copying the env and copy more
-        // update files than not enough.
-        //
-        // Note that there cannot be any update files deleted between those
-        // two read operations as the task processing is synchronous.
-
-        // 4.1 First copy the LMDB env of the index-scheduler
-        progress.update_progress(SnapshotCreationProgress::SnapshotTheIndexScheduler);
-        builder.add_env_to_archive(&self.env)?;
-        if must_stop_processing.get() {
-            return abort_no_index;
-        }
-
-        // 4.2 Create a read transaction on the index-scheduler
-        let rtxn = self.env.read_txn()?;
-
-        // 4.3 Only copy the update files of the enqueued tasks
-        progress.update_progress(SnapshotCreationProgress::SnapshotTheUpdateFiles);
-        builder.add_dir_to_archive(self.queue.file_store.path().to_path_buf())?;
-        let enqueued = self.queue.tasks.get_status(&rtxn, Status::Enqueued)?;
-        let (atomic, update_file_progress) = AtomicUpdateFileStep::new(enqueued.len() as u32);
-        progress.update_progress(update_file_progress);
-        for task_id in enqueued {
-            if must_stop_processing.get() {
-                return abort_no_index;
-            }
-            let task =
-                self.queue.tasks.get_task(&rtxn, task_id)?.ok_or(Error::CorruptedTaskQueue)?;
-            if let Some(content_uuid) = task.content_uuid() {
-                let src = self.queue.file_store.get_update_path(content_uuid);
-                builder.add_file_to_archive(src)?;
-            }
-            atomic.fetch_add(1, Ordering::Relaxed);
-        }
-
-        // 5. Snapshot every index
-        progress.update_progress(SnapshotCreationProgress::SnapshotTheIndexes);
-        builder.add_dir_to_archive(self.index_mapper.base_path().to_path_buf())?;
-        let index_mapping = self.index_mapper.index_mapping;
-        let nb_indexes = index_mapping.len(&rtxn)? as u32;
-
-        for (i, result) in index_mapping.iter(&rtxn)?.enumerate() {
-            let (name, _) = result?;
-            let abort_index = || {
-                Err(Error::from_milli(
-                    InternalError::AbortedIndexation.into(),
-                    Some(name.to_string()), // defer the `to_string`
-                ))
-            };
-
-            if must_stop_processing.get() {
-                return abort_index();
-            }
-
-            progress.update_progress(VariableNameStep::<SnapshotCreationProgress>::new(
-                name, i as u32, nb_indexes,
-            ));
-            let index = self.index_mapper.index(&rtxn, name)?;
-            builder.add_env_to_archive(index.raw_env())?;
-        }
-
-        drop(rtxn);
-
-        if must_stop_processing.get() {
-            return abort_no_index;
-        }
-
-        // 6. Snapshot the auth LMDB env
-        progress.update_progress(SnapshotCreationProgress::SnapshotTheApiKeys);
-        builder.add_env_to_archive(&self.scheduler.auth_env)?;
-
-        // 7. Finalize the tarball
-        progress.update_progress(SnapshotCreationProgress::CreateTheTarball);
-        let file =
-            builder.finish(&self.scheduler.snapshots_path.join(format!("{db_name}.snapshot")))?;
-
-        // 8. Change the permission to make the snapshot readonly
-        let mut permissions = file.metadata()?.permissions();
-        permissions.set_readonly(true);
-        #[cfg(unix)]
-        {
-            use std::os::unix::fs::PermissionsExt;
-            #[allow(clippy::non_octal_unix_permissions)]
-            //                     rwxrwxrwx
-            permissions.set_mode(0b100100100);
-        }
-
-        file.set_permissions(permissions)?;
-
-        for task in &mut tasks {
-            task.status = Status::Succeeded;
-        }
-
-        Ok(tasks)
-    }
 }

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

@@ -6,7 +6,7 @@ source: crates/index-scheduler/src/scheduler/test_failure.rs
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, batch_uid: 0, status: succeeded, details: { from: (1, 12, 0), to: (1, 22, 1) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
+0 {uid: 0, batch_uid: 0, status: succeeded, details: { from: (1, 12, 0), to: (1, 24, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
 1 {uid: 1, batch_uid: 1, status: succeeded, details: { primary_key: Some("mouse"), old_new_uid: None, new_index_uid: None }, kind: IndexCreation { index_uid: "catto", primary_key: Some("mouse") }}
 2 {uid: 2, batch_uid: 2, status: succeeded, details: { primary_key: Some("bone"), old_new_uid: None, new_index_uid: None }, 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"), old_new_uid: None, new_index_uid: None }, kind: IndexCreation { index_uid: "doggo", primary_key: Some("bone") }}
@@ -57,7 +57,7 @@ girafo: { number_of_documents: 0, field_distribution: {} }
 [timestamp] [4,]
 ----------------------------------------------------------------------
 ### All Batches:
-0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.22.1"}, stats: {"totalNbTasks":1,"status":{"succeeded":1},"types":{"upgradeDatabase":1},"indexUids":{}}, stop reason: "stopped after the last task of type `upgradeDatabase` because they cannot be batched with tasks of any other type.", }
+0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.24.0"}, stats: {"totalNbTasks":1,"status":{"succeeded":1},"types":{"upgradeDatabase":1},"indexUids":{}}, stop reason: "stopped after the last task of type `upgradeDatabase` because they cannot be batched with tasks of any other type.", }
 1 {uid: 1, details: {"primaryKey":"mouse"}, stats: {"totalNbTasks":1,"status":{"succeeded":1},"types":{"indexCreation":1},"indexUids":{"catto":1}}, stop reason: "created batch containing only task with id 1 of type `indexCreation` that cannot be batched with any other task.", }
 2 {uid: 2, details: {"primaryKey":"bone"}, stats: {"totalNbTasks":1,"status":{"succeeded":1},"types":{"indexCreation":1},"indexUids":{"doggo":1}}, stop reason: "created batch containing only task with id 2 of type `indexCreation` that cannot be batched with any other task.", }
 3 {uid: 3, details: {"primaryKey":"bone"}, stats: {"totalNbTasks":1,"status":{"failed":1},"types":{"indexCreation":1},"indexUids":{"doggo":1}}, stop reason: "created batch containing only task with id 3 of type `indexCreation` that cannot be batched with any other task.", }

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

@@ -6,7 +6,7 @@ source: crates/index-scheduler/src/scheduler/test_failure.rs
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, status: enqueued, details: { from: (1, 12, 0), to: (1, 22, 1) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
+0 {uid: 0, status: enqueued, details: { from: (1, 12, 0), to: (1, 24, 0) }, 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

@@ -6,7 +6,7 @@ source: crates/index-scheduler/src/scheduler/test_failure.rs
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, status: enqueued, details: { from: (1, 12, 0), to: (1, 22, 1) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
+0 {uid: 0, status: enqueued, details: { from: (1, 12, 0), to: (1, 24, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
 1 {uid: 1, status: enqueued, details: { primary_key: Some("mouse"), old_new_uid: None, new_index_uid: None }, 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

@@ -6,7 +6,7 @@ source: crates/index-scheduler/src/scheduler/test_failure.rs
 []
 ----------------------------------------------------------------------
 ### 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, 22, 1) }, 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, 24, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
 1 {uid: 1, status: enqueued, details: { primary_key: Some("mouse"), old_new_uid: None, new_index_uid: None }, kind: IndexCreation { index_uid: "catto", primary_key: Some("mouse") }}
 ----------------------------------------------------------------------
 ### Status:
@@ -37,7 +37,7 @@ catto [1,]
 [timestamp] [0,]
 ----------------------------------------------------------------------
 ### All Batches:
-0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.22.1"}, stats: {"totalNbTasks":1,"status":{"failed":1},"types":{"upgradeDatabase":1},"indexUids":{}}, stop reason: "stopped after the last task of type `upgradeDatabase` because they cannot be batched with tasks of any other type.", }
+0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.24.0"}, stats: {"totalNbTasks":1,"status":{"failed":1},"types":{"upgradeDatabase":1},"indexUids":{}}, stop reason: "stopped after the last task of type `upgradeDatabase` because they cannot be batched with tasks of any other type.", }
 ----------------------------------------------------------------------
 ### Batch to tasks mapping:
 0 [0,]

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

@@ -6,7 +6,7 @@ source: crates/index-scheduler/src/scheduler/test_failure.rs
 []
 ----------------------------------------------------------------------
 ### 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, 22, 1) }, 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, 24, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
 1 {uid: 1, status: enqueued, details: { primary_key: Some("mouse"), old_new_uid: None, new_index_uid: None }, kind: IndexCreation { index_uid: "catto", primary_key: Some("mouse") }}
 2 {uid: 2, status: enqueued, details: { primary_key: Some("bone"), old_new_uid: None, new_index_uid: None }, kind: IndexCreation { index_uid: "doggo", primary_key: Some("bone") }}
 ----------------------------------------------------------------------
@@ -40,7 +40,7 @@ doggo [2,]
 [timestamp] [0,]
 ----------------------------------------------------------------------
 ### All Batches:
-0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.22.1"}, stats: {"totalNbTasks":1,"status":{"failed":1},"types":{"upgradeDatabase":1},"indexUids":{}}, stop reason: "stopped after the last task of type `upgradeDatabase` because they cannot be batched with tasks of any other type.", }
+0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.24.0"}, stats: {"totalNbTasks":1,"status":{"failed":1},"types":{"upgradeDatabase":1},"indexUids":{}}, stop reason: "stopped after the last task of type `upgradeDatabase` because they cannot be batched with tasks of any other type.", }
 ----------------------------------------------------------------------
 ### Batch to tasks mapping:
 0 [0,]

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

@@ -6,7 +6,7 @@ source: crates/index-scheduler/src/scheduler/test_failure.rs
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, batch_uid: 0, status: succeeded, details: { from: (1, 12, 0), to: (1, 22, 1) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
+0 {uid: 0, batch_uid: 0, status: succeeded, details: { from: (1, 12, 0), to: (1, 24, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
 1 {uid: 1, status: enqueued, details: { primary_key: Some("mouse"), old_new_uid: None, new_index_uid: None }, kind: IndexCreation { index_uid: "catto", primary_key: Some("mouse") }}
 2 {uid: 2, status: enqueued, details: { primary_key: Some("bone"), old_new_uid: None, new_index_uid: None }, kind: IndexCreation { index_uid: "doggo", primary_key: Some("bone") }}
 3 {uid: 3, status: enqueued, details: { primary_key: Some("bone"), old_new_uid: None, new_index_uid: None }, kind: IndexCreation { index_uid: "doggo", primary_key: Some("bone") }}
@@ -43,7 +43,7 @@ doggo [2,3,]
 [timestamp] [0,]
 ----------------------------------------------------------------------
 ### All Batches:
-0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.22.1"}, stats: {"totalNbTasks":1,"status":{"succeeded":1},"types":{"upgradeDatabase":1},"indexUids":{}}, stop reason: "stopped after the last task of type `upgradeDatabase` because they cannot be batched with tasks of any other type.", }
+0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.24.0"}, stats: {"totalNbTasks":1,"status":{"succeeded":1},"types":{"upgradeDatabase":1},"indexUids":{}}, stop reason: "stopped after the last task of type `upgradeDatabase` because they cannot be batched with tasks of any other type.", }
 ----------------------------------------------------------------------
 ### Batch to tasks mapping:
 0 [0,]

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

@@ -722,7 +722,7 @@ fn basic_get_stats() {
     let kind = index_creation_task("whalo", "fish");
     let _task = index_scheduler.register(kind, None, false).unwrap();
 
-    snapshot!(json_string!(index_scheduler.get_stats().unwrap()), @r#"
+    snapshot!(json_string!(index_scheduler.get_stats().unwrap()), @r###"
     {
       "indexes": {
         "catto": 1,
@@ -742,6 +742,7 @@ fn basic_get_stats() {
         "documentEdition": 0,
         "dumpCreation": 0,
         "export": 0,
+        "indexCompaction": 0,
         "indexCreation": 3,
         "indexDeletion": 0,
         "indexSwap": 0,
@@ -753,10 +754,10 @@ fn basic_get_stats() {
         "upgradeDatabase": 0
       }
     }
-    "#);
+    "###);
 
     handle.advance_till([Start, BatchCreated]);
-    snapshot!(json_string!(index_scheduler.get_stats().unwrap()), @r#"
+    snapshot!(json_string!(index_scheduler.get_stats().unwrap()), @r###"
     {
       "indexes": {
         "catto": 1,
@@ -776,6 +777,7 @@ fn basic_get_stats() {
         "documentEdition": 0,
         "dumpCreation": 0,
         "export": 0,
+        "indexCompaction": 0,
         "indexCreation": 3,
         "indexDeletion": 0,
         "indexSwap": 0,
@@ -787,7 +789,7 @@ fn basic_get_stats() {
         "upgradeDatabase": 0
       }
     }
-    "#);
+    "###);
 
     handle.advance_till([
         InsideProcessBatch,
@@ -797,7 +799,7 @@ fn basic_get_stats() {
         Start,
         BatchCreated,
     ]);
-    snapshot!(json_string!(index_scheduler.get_stats().unwrap()), @r#"
+    snapshot!(json_string!(index_scheduler.get_stats().unwrap()), @r###"
     {
       "indexes": {
         "catto": 1,
@@ -817,6 +819,7 @@ fn basic_get_stats() {
         "documentEdition": 0,
         "dumpCreation": 0,
         "export": 0,
+        "indexCompaction": 0,
         "indexCreation": 3,
         "indexDeletion": 0,
         "indexSwap": 0,
@@ -828,7 +831,7 @@ fn basic_get_stats() {
         "upgradeDatabase": 0
       }
     }
-    "#);
+    "###);
 
     // now we make one more batch, the started_at field of the new tasks will be past `second_start_time`
     handle.advance_till([
@@ -839,7 +842,7 @@ fn basic_get_stats() {
         Start,
         BatchCreated,
     ]);
-    snapshot!(json_string!(index_scheduler.get_stats().unwrap()), @r#"
+    snapshot!(json_string!(index_scheduler.get_stats().unwrap()), @r###"
     {
       "indexes": {
         "catto": 1,
@@ -859,6 +862,7 @@ fn basic_get_stats() {
         "documentEdition": 0,
         "dumpCreation": 0,
         "export": 0,
+        "indexCompaction": 0,
         "indexCreation": 3,
         "indexDeletion": 0,
         "indexSwap": 0,
@@ -870,7 +874,7 @@ fn basic_get_stats() {
         "upgradeDatabase": 0
       }
     }
-    "#);
+    "###);
 }
 
 #[test]

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

@@ -46,6 +46,8 @@ pub fn upgrade_index_scheduler(
         (1, 20, _) => 0,
         (1, 21, _) => 0,
         (1, 22, _) => 0,
+        (1, 23, _) => 0,
+        (1, 24, _) => 0,
         (major, minor, patch) => {
             if major > current_major
                 || (major == current_major && minor > current_minor)

crates/index-scheduler/src/utils.rs

@@ -256,14 +256,15 @@ pub fn swap_index_uid_in_task(task: &mut Task, swap: (&str, &str)) {
     use KindWithContent as K;
     let mut index_uids = vec![];
     match &mut task.kind {
-        K::DocumentAdditionOrUpdate { index_uid, .. } => index_uids.push(index_uid),
-        K::DocumentEdition { index_uid, .. } => index_uids.push(index_uid),
-        K::DocumentDeletion { index_uid, .. } => index_uids.push(index_uid),
-        K::DocumentDeletionByFilter { index_uid, .. } => index_uids.push(index_uid),
-        K::DocumentClear { index_uid } => index_uids.push(index_uid),
-        K::SettingsUpdate { index_uid, .. } => index_uids.push(index_uid),
-        K::IndexDeletion { index_uid } => index_uids.push(index_uid),
-        K::IndexCreation { index_uid, .. } => index_uids.push(index_uid),
+        K::DocumentAdditionOrUpdate { index_uid, .. }
+        | K::DocumentEdition { index_uid, .. }
+        | K::DocumentDeletion { index_uid, .. }
+        | K::DocumentDeletionByFilter { index_uid, .. }
+        | K::DocumentClear { index_uid }
+        | K::SettingsUpdate { index_uid, .. }
+        | K::IndexDeletion { index_uid }
+        | K::IndexCreation { index_uid, .. }
+        | K::IndexCompaction { index_uid, .. } => index_uids.push(index_uid),
         K::IndexUpdate { index_uid, new_index_uid, .. } => {
             index_uids.push(index_uid);
             if let Some(new_uid) = new_index_uid {
@@ -618,6 +619,13 @@ impl crate::IndexScheduler {
                     Details::UpgradeDatabase { from: _, to: _ } => {
                         assert_eq!(kind.as_kind(), Kind::UpgradeDatabase);
                     }
+                    Details::IndexCompaction {
+                        index_uid: _,
+                        pre_compaction_size: _,
+                        post_compaction_size: _,
+                    } => {
+                        assert_eq!(kind.as_kind(), Kind::IndexCompaction);
+                    }
                 }
             }
 

crates/meilisearch-auth/src/store.rs

@@ -109,6 +109,7 @@ impl HeedAuthStore {
                             Action::IndexesGet,
                             Action::IndexesUpdate,
                             Action::IndexesSwap,
+                            Action::IndexesCompact,
                         ]
                         .iter(),
                     );

crates/meilisearch-types/src/compression.rs

@@ -1,13 +1,11 @@
 use std::fs::{create_dir_all, File};
 use std::io::Write;
-use std::path::{Path, PathBuf};
+use std::path::Path;
 
 use flate2::read::GzDecoder;
 use flate2::write::GzEncoder;
 use flate2::Compression;
-use milli::heed::Env;
 use tar::{Archive, Builder};
-use tempfile::NamedTempFile;
 
 pub fn to_tar_gz(src: impl AsRef<Path>, dest: impl AsRef<Path>) -> anyhow::Result<()> {
     let mut f = File::create(dest)?;
@@ -28,112 +26,3 @@ pub fn from_tar_gz(src: impl AsRef<Path>, dest: impl AsRef<Path>) -> anyhow::Res
     ar.unpack(&dest)?;
     Ok(())
 }
-
-pub struct PipedArchiveBuilder {
-    base_path: PathBuf,
-    tar_encoder: tar::Builder<GzEncoder<NamedTempFile>>,
-}
-
-impl PipedArchiveBuilder {
-    pub fn new(dest_dir: PathBuf, base_path: PathBuf) -> anyhow::Result<Self> {
-        let temp_archive = tempfile::NamedTempFile::new_in(&dest_dir)?;
-
-        let gz_encoder = GzEncoder::new(temp_archive, Compression::default());
-        let mut tar_encoder = Builder::new(gz_encoder);
-        let base_path_in_archive = PathInArchive::from_absolute_and_base(&base_path, &base_path);
-        tar_encoder.append_dir(base_path_in_archive.as_path(), &base_path)?;
-
-        Ok(Self { base_path, tar_encoder })
-    }
-
-    /// Add a heed environment to the archive.
-    ///
-    /// # Errors
-    ///
-    /// - Errors originating with that thread:
-    ///     - Heed errors, if taking a write transaction fails
-    ///     - If the copy of the environment fails.
-    ///     - If there is an I/O error opening the database at the environment's path.
-    /// - Errors originating with another thread:
-    ///     - If the cancellation thread panicked or otherwise dropped its receiver.
-    ///     - If the processing thread panicked or otherwise dropped its receiver.
-    pub fn add_env_to_archive<T>(&mut self, env: &Env<T>) -> anyhow::Result<()> {
-        let path = env.path().to_path_buf();
-        // make sure that the environment cannot change while it is being added to the archive,
-        // as any concurrent change would corrupt the copy.
-        let env_wtxn = env.write_txn()?;
-
-        let dir_path_in_archive = PathInArchive::from_absolute_and_base(&path, &self.base_path);
-
-        self.tar_encoder.append_dir(dir_path_in_archive.as_path(), &path)?;
-
-        let path = path.join("data.mdb");
-        let path_in_archive = PathInArchive::from_absolute_and_base(&path, &self.base_path);
-
-        self.tar_encoder.append_path_with_name(&path, path_in_archive.as_path())?;
-
-        // no change we might want to commit
-        env_wtxn.abort();
-        Ok(())
-    }
-
-    /// Add a file to the archive
-    ///
-    /// # Errors
-    ///
-    /// - If the processing thread panicked or otherwise dropped its receiver.
-    pub fn add_file_to_archive(&mut self, path: PathBuf) -> anyhow::Result<()> {
-        let path_in_archive = PathInArchive::from_absolute_and_base(&path, &self.base_path);
-        self.tar_encoder.append_path_with_name(&path, path_in_archive.as_path())?;
-        Ok(())
-    }
-
-    /// Add a directory name (**without its contents**) to the archive.
-    ///
-    /// # Errors
-    ///
-    /// - If the processing thread panicked or otherwise dropped its receiver.
-    pub fn add_dir_to_archive(&mut self, path: PathBuf) -> anyhow::Result<()> {
-        let path_in_archive = PathInArchive::from_absolute_and_base(&path, &self.base_path);
-
-        self.tar_encoder.append_dir(path_in_archive.as_path(), &path)?;
-        Ok(())
-    }
-
-    /// Finalize the archive and persists it to disk.
-    ///
-    /// # Errors
-    ///
-    /// - Originating with the current thread:
-    ///     - If persisting the archive fails
-    /// - Originating with another thread:
-    ///     - If the cancellation thread panicked.
-    ///     - If the processing thread panicked or otherwise terminated in error.
-    pub fn finish(self, dest_path: &Path) -> anyhow::Result<File> {
-        let gz_encoder = self.tar_encoder.into_inner()?;
-        let mut temp_archive = gz_encoder.finish()?;
-        temp_archive.flush()?;
-
-        let archive = temp_archive.persist(dest_path)?;
-        Ok(archive)
-    }
-}
-
-struct PathInArchive(PathBuf);
-
-impl PathInArchive {
-    pub fn from_absolute_and_base(absolute: &Path, base: &Path) -> Self {
-        /// FIXME
-        let canonical = absolute.canonicalize().unwrap();
-        let relative = match canonical.strip_prefix(base) {
-            Ok(stripped) => Path::new(&".").join(stripped),
-            Err(_) => absolute.to_path_buf(),
-        };
-
-        Self(relative)
-    }
-
-    pub fn as_path(&self) -> &Path {
-        self.0.as_path()
-    }
-}

crates/meilisearch-types/src/keys.rs

@@ -380,6 +380,9 @@ pub enum Action {
     #[serde(rename = "webhooks.*")]
     #[deserr(rename = "webhooks.*")]
     WebhooksAll,
+    #[serde(rename = "indexes.compact")]
+    #[deserr(rename = "indexes.compact")]
+    IndexesCompact,
 }
 
 impl Action {
@@ -398,6 +401,7 @@ impl Action {
             INDEXES_UPDATE => Some(Self::IndexesUpdate),
             INDEXES_DELETE => Some(Self::IndexesDelete),
             INDEXES_SWAP => Some(Self::IndexesSwap),
+            INDEXES_COMPACT => Some(Self::IndexesCompact),
             TASKS_ALL => Some(Self::TasksAll),
             TASKS_CANCEL => Some(Self::TasksCancel),
             TASKS_DELETE => Some(Self::TasksDelete),
@@ -462,6 +466,7 @@ impl Action {
             IndexesUpdate => false,
             IndexesDelete => false,
             IndexesSwap => false,
+            IndexesCompact => false,
             TasksCancel => false,
             TasksDelete => false,
             TasksGet => true,
@@ -513,6 +518,7 @@ pub mod actions {
     pub const INDEXES_UPDATE: u8 = IndexesUpdate.repr();
     pub const INDEXES_DELETE: u8 = IndexesDelete.repr();
     pub const INDEXES_SWAP: u8 = IndexesSwap.repr();
+    pub const INDEXES_COMPACT: u8 = IndexesCompact.repr();
     pub const TASKS_ALL: u8 = TasksAll.repr();
     pub const TASKS_CANCEL: u8 = TasksCancel.repr();
     pub const TASKS_DELETE: u8 = TasksDelete.repr();
@@ -614,6 +620,7 @@ pub(crate) mod test {
         assert!(WebhooksDelete.repr() == 47 && WEBHOOKS_DELETE == 47);
         assert!(WebhooksCreate.repr() == 48 && WEBHOOKS_CREATE == 48);
         assert!(WebhooksAll.repr() == 49 && WEBHOOKS_ALL == 49);
+        assert!(IndexesCompact.repr() == 50 && INDEXES_COMPACT == 50);
     }
 
     #[test]

crates/meilisearch-types/src/task_view.rs

@@ -142,6 +142,11 @@ pub struct DetailsView {
     pub old_index_uid: Option<String>,
     #[serde(skip_serializing_if = "Option::is_none")]
     pub new_index_uid: Option<String>,
+    // index compaction
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub pre_compaction_size: Option<String>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub post_compaction_size: Option<String>,
 }
 
 impl DetailsView {
@@ -314,6 +319,24 @@ impl DetailsView {
                 // We should never be able to batch multiple renames at the same time.
                 (Some(left), Some(_right)) => Some(left),
             },
+            pre_compaction_size: match (
+                self.pre_compaction_size.clone(),
+                other.pre_compaction_size.clone(),
+            ) {
+                (None, None) => None,
+                (None, Some(size)) | (Some(size), None) => Some(size),
+                // We should never be able to batch multiple compactions at the same time.
+                (Some(left), Some(_right)) => Some(left),
+            },
+            post_compaction_size: match (
+                self.post_compaction_size.clone(),
+                other.post_compaction_size.clone(),
+            ) {
+                (None, None) => None,
+                (None, Some(size)) | (Some(size), None) => Some(size),
+                // We should never be able to batch multiple compactions at the same time.
+                (Some(left), Some(_right)) => Some(left),
+            },
         }
     }
 }
@@ -415,6 +438,15 @@ impl From<Details> for DetailsView {
                 upgrade_to: Some(format!("v{}.{}.{}", to.0, to.1, to.2)),
                 ..Default::default()
             },
+            Details::IndexCompaction { pre_compaction_size, post_compaction_size, .. } => {
+                DetailsView {
+                    pre_compaction_size: pre_compaction_size
+                        .map(|size| size.get_appropriate_unit(UnitType::Both).to_string()),
+                    post_compaction_size: post_compaction_size
+                        .map(|size| size.get_appropriate_unit(UnitType::Both).to_string()),
+                    ..Default::default()
+                }
+            }
         }
     }
 }

crates/meilisearch-types/src/tasks.rs

@@ -67,7 +67,8 @@ impl Task {
             | SettingsUpdate { index_uid, .. }
             | IndexCreation { index_uid, .. }
             | IndexUpdate { index_uid, .. }
-            | IndexDeletion { index_uid } => Some(index_uid),
+            | IndexDeletion { index_uid }
+            | IndexCompaction { index_uid } => Some(index_uid),
         }
     }
 
@@ -94,7 +95,8 @@ impl Task {
             | KindWithContent::DumpCreation { .. }
             | KindWithContent::SnapshotCreation
             | KindWithContent::Export { .. }
-            | KindWithContent::UpgradeDatabase { .. } => None,
+            | KindWithContent::UpgradeDatabase { .. }
+            | KindWithContent::IndexCompaction { .. } => None,
         }
     }
 }
@@ -170,6 +172,9 @@ pub enum KindWithContent {
     UpgradeDatabase {
         from: (u32, u32, u32),
     },
+    IndexCompaction {
+        index_uid: String,
+    },
 }
 
 #[derive(Debug, Clone, PartialEq, Eq, Serialize, Deserialize, ToSchema)]
@@ -206,6 +211,7 @@ impl KindWithContent {
             KindWithContent::SnapshotCreation => Kind::SnapshotCreation,
             KindWithContent::Export { .. } => Kind::Export,
             KindWithContent::UpgradeDatabase { .. } => Kind::UpgradeDatabase,
+            KindWithContent::IndexCompaction { .. } => Kind::IndexCompaction,
         }
     }
 
@@ -226,7 +232,8 @@ impl KindWithContent {
             | DocumentClear { index_uid }
             | SettingsUpdate { index_uid, .. }
             | IndexCreation { index_uid, .. }
-            | IndexDeletion { index_uid } => vec![index_uid],
+            | IndexDeletion { index_uid }
+            | IndexCompaction { index_uid } => vec![index_uid],
             IndexUpdate { index_uid, new_index_uid, .. } => {
                 let mut indexes = vec![index_uid.as_str()];
                 if let Some(new_uid) = new_index_uid {
@@ -325,6 +332,11 @@ impl KindWithContent {
                     versioning::VERSION_PATCH,
                 ),
             }),
+            KindWithContent::IndexCompaction { index_uid } => Some(Details::IndexCompaction {
+                index_uid: index_uid.clone(),
+                pre_compaction_size: None,
+                post_compaction_size: None,
+            }),
         }
     }
 
@@ -407,6 +419,11 @@ impl KindWithContent {
                     versioning::VERSION_PATCH,
                 ),
             }),
+            KindWithContent::IndexCompaction { index_uid } => Some(Details::IndexCompaction {
+                index_uid: index_uid.clone(),
+                pre_compaction_size: None,
+                post_compaction_size: None,
+            }),
         }
     }
 }
@@ -469,6 +486,11 @@ impl From<&KindWithContent> for Option<Details> {
                     versioning::VERSION_PATCH,
                 ),
             }),
+            KindWithContent::IndexCompaction { index_uid } => Some(Details::IndexCompaction {
+                index_uid: index_uid.clone(),
+                pre_compaction_size: None,
+                post_compaction_size: None,
+            }),
         }
     }
 }
@@ -579,6 +601,7 @@ pub enum Kind {
     SnapshotCreation,
     Export,
     UpgradeDatabase,
+    IndexCompaction,
 }
 
 impl Kind {
@@ -590,7 +613,8 @@ impl Kind {
             | Kind::SettingsUpdate
             | Kind::IndexCreation
             | Kind::IndexDeletion
-            | Kind::IndexUpdate => true,
+            | Kind::IndexUpdate
+            | Kind::IndexCompaction => true,
             Kind::IndexSwap
             | Kind::TaskCancelation
             | Kind::TaskDeletion
@@ -618,6 +642,7 @@ impl Display for Kind {
             Kind::SnapshotCreation => write!(f, "snapshotCreation"),
             Kind::Export => write!(f, "export"),
             Kind::UpgradeDatabase => write!(f, "upgradeDatabase"),
+            Kind::IndexCompaction => write!(f, "indexCompaction"),
         }
     }
 }
@@ -653,6 +678,8 @@ impl FromStr for Kind {
             Ok(Kind::Export)
         } else if kind.eq_ignore_ascii_case("upgradeDatabase") {
             Ok(Kind::UpgradeDatabase)
+        } else if kind.eq_ignore_ascii_case("indexCompaction") {
+            Ok(Kind::IndexCompaction)
         } else {
             Err(ParseTaskKindError(kind.to_owned()))
         }
@@ -738,6 +765,11 @@ pub enum Details {
         from: (u32, u32, u32),
         to: (u32, u32, u32),
     },
+    IndexCompaction {
+        index_uid: String,
+        pre_compaction_size: Option<Byte>,
+        post_compaction_size: Option<Byte>,
+    },
 }
 
 #[derive(Debug, PartialEq, Clone, Serialize, Deserialize, ToSchema)]
@@ -800,6 +832,10 @@ impl Details {
             Self::ClearAll { deleted_documents } => *deleted_documents = Some(0),
             Self::TaskCancelation { canceled_tasks, .. } => *canceled_tasks = Some(0),
             Self::TaskDeletion { deleted_tasks, .. } => *deleted_tasks = Some(0),
+            Self::IndexCompaction { pre_compaction_size, post_compaction_size, .. } => {
+                *pre_compaction_size = None;
+                *post_compaction_size = None;
+            }
             Self::SettingsUpdate { .. }
             | Self::IndexInfo { .. }
             | Self::Dump { .. }

crates/meilisearch/Cargo.toml

@@ -91,7 +91,7 @@ time = { version = "0.3.41", features = [
 ] }
 tokio = { version = "1.45.1", features = ["full"] }
 toml = "0.8.23"
-uuid = { version = "1.17.0", features = ["serde", "v4"] }
+uuid = { version = "1.18.0", features = ["serde", "v4", "v7"] }
 serde_urlencoded = "0.7.1"
 termcolor = "1.4.1"
 url = { version = "2.5.4", features = ["serde"] }
@@ -117,7 +117,7 @@ secrecy = "0.10.3"
 actix-web-lab = { version = "0.24.1", default-features = false }
 urlencoding = "2.1.3"
 backoff = { version = "0.4.0", features = ["tokio"] }
-
+humantime = { version = "2.3.0", default-features = false }
 
 [dev-dependencies]
 actix-rt = "2.10.0"

crates/meilisearch/src/analytics/segment_analytics.rs

@@ -205,6 +205,8 @@ struct Infos {
     experimental_no_snapshot_compaction: bool,
     experimental_no_edition_2024_for_dumps: bool,
     experimental_no_edition_2024_for_settings: bool,
+    experimental_no_edition_2024_for_prefix_post_processing: bool,
+    experimental_no_edition_2024_for_facet_post_processing: bool,
     experimental_vector_store_setting: bool,
     gpu_enabled: bool,
     db_path: bool,
@@ -296,6 +298,8 @@ impl Infos {
             skip_index_budget: _,
             experimental_no_edition_2024_for_settings,
             experimental_no_edition_2024_for_dumps,
+            experimental_no_edition_2024_for_prefix_post_processing,
+            experimental_no_edition_2024_for_facet_post_processing,
         } = indexer_options;
 
         let RuntimeTogglableFeatures {
@@ -365,6 +369,8 @@ impl Infos {
             ssl_resumption,
             ssl_tickets,
             experimental_no_edition_2024_for_settings,
+            experimental_no_edition_2024_for_prefix_post_processing,
+            experimental_no_edition_2024_for_facet_post_processing,
         }
     }
 }

crates/meilisearch/src/metrics.rs

@@ -1,7 +1,8 @@
 use lazy_static::lazy_static;
 use prometheus::{
-    opts, register_gauge, register_histogram_vec, register_int_counter_vec, register_int_gauge,
-    register_int_gauge_vec, Gauge, HistogramVec, IntCounterVec, IntGauge, IntGaugeVec,
+    opts, register_gauge, register_gauge_vec, register_histogram_vec, register_int_counter_vec,
+    register_int_gauge, register_int_gauge_vec, Gauge, GaugeVec, HistogramVec, IntCounterVec,
+    IntGauge, IntGaugeVec,
 };
 
 lazy_static! {
@@ -73,6 +74,20 @@ lazy_static! {
         &["kind", "value"]
     )
     .expect("Can't create a metric");
+    pub static ref MEILISEARCH_BATCH_RUNNING_PROGRESS_TRACE: GaugeVec = register_gauge_vec!(
+        opts!("meilisearch_batch_running_progress_trace", "The currently running progress trace"),
+        &["batch_uid", "step_name"]
+    )
+    .expect("Can't create a metric");
+    pub static ref MEILISEARCH_LAST_FINISHED_BATCHES_PROGRESS_TRACE_MS: IntGaugeVec =
+        register_int_gauge_vec!(
+            opts!(
+                "meilisearch_last_finished_batches_progress_trace_ms",
+                "The last few batches progress trace in milliseconds"
+            ),
+            &["batch_uid", "step_name"]
+        )
+        .expect("Can't create a metric");
     pub static ref MEILISEARCH_LAST_UPDATE: IntGauge =
         register_int_gauge!(opts!("meilisearch_last_update", "Meilisearch Last Update"))
             .expect("Can't create a metric");

crates/meilisearch/src/option.rs

@@ -55,6 +55,10 @@ const MEILI_EXPERIMENTAL_ENABLE_LOGS_ROUTE: &str = "MEILI_EXPERIMENTAL_ENABLE_LO
 const MEILI_EXPERIMENTAL_CONTAINS_FILTER: &str = "MEILI_EXPERIMENTAL_CONTAINS_FILTER";
 const MEILI_EXPERIMENTAL_NO_EDITION_2024_FOR_SETTINGS: &str =
     "MEILI_EXPERIMENTAL_NO_EDITION_2024_FOR_SETTINGS";
+const MEILI_EXPERIMENTAL_NO_EDITION_2024_FOR_FACET_POST_PROCESSING: &str =
+    "MEILI_EXPERIMENTAL_NO_EDITION_2024_FOR_FACET_POST_PROCESSING";
+const MEILI_EXPERIMENTAL_NO_EDITION_2024_FOR_PREFIX_POST_PROCESSING: &str =
+    "MEILI_EXPERIMENTAL_NO_EDITION_2024_FOR_PREFIX_POST_PROCESSING";
 const MEILI_EXPERIMENTAL_ENABLE_METRICS: &str = "MEILI_EXPERIMENTAL_ENABLE_METRICS";
 const MEILI_EXPERIMENTAL_SEARCH_QUEUE_SIZE: &str = "MEILI_EXPERIMENTAL_SEARCH_QUEUE_SIZE";
 const MEILI_EXPERIMENTAL_DROP_SEARCH_AFTER: &str = "MEILI_EXPERIMENTAL_DROP_SEARCH_AFTER";
@@ -772,6 +776,22 @@ pub struct IndexerOpts {
     #[clap(long, env = MEILI_EXPERIMENTAL_NO_EDITION_2024_FOR_DUMPS)]
     #[serde(default)]
     pub experimental_no_edition_2024_for_dumps: bool,
+
+    /// Experimental no edition 2024 to compute prefixes. For more information,
+    /// see: <https://github.com/orgs/meilisearch/discussions/862>
+    ///
+    /// Enables the experimental no edition 2024 to compute prefixes.
+    #[clap(long, env = MEILI_EXPERIMENTAL_NO_EDITION_2024_FOR_PREFIX_POST_PROCESSING)]
+    #[serde(default)]
+    pub experimental_no_edition_2024_for_prefix_post_processing: bool,
+
+    /// Experimental no edition 2024 to compute facets. For more information,
+    /// see: <https://github.com/orgs/meilisearch/discussions/862>
+    ///
+    /// Enables the experimental no edition 2024 to compute facets.
+    #[clap(long, env = MEILI_EXPERIMENTAL_NO_EDITION_2024_FOR_FACET_POST_PROCESSING)]
+    #[serde(default)]
+    pub experimental_no_edition_2024_for_facet_post_processing: bool,
 }
 
 impl IndexerOpts {
@@ -783,6 +803,8 @@ impl IndexerOpts {
             skip_index_budget: _,
             experimental_no_edition_2024_for_settings,
             experimental_no_edition_2024_for_dumps,
+            experimental_no_edition_2024_for_prefix_post_processing,
+            experimental_no_edition_2024_for_facet_post_processing,
         } = self;
         if let Some(max_indexing_memory) = max_indexing_memory.0 {
             export_to_env_if_not_present(
@@ -808,6 +830,18 @@ impl IndexerOpts {
                 experimental_no_edition_2024_for_dumps.to_string(),
             );
         }
+        if experimental_no_edition_2024_for_prefix_post_processing {
+            export_to_env_if_not_present(
+                MEILI_EXPERIMENTAL_NO_EDITION_2024_FOR_PREFIX_POST_PROCESSING,
+                experimental_no_edition_2024_for_prefix_post_processing.to_string(),
+            );
+        }
+        if experimental_no_edition_2024_for_facet_post_processing {
+            export_to_env_if_not_present(
+                MEILI_EXPERIMENTAL_NO_EDITION_2024_FOR_FACET_POST_PROCESSING,
+                experimental_no_edition_2024_for_facet_post_processing.to_string(),
+            );
+        }
     }
 }
 
@@ -833,6 +867,10 @@ impl TryFrom<&IndexerOpts> for IndexerConfig {
             chunk_compression_level: Default::default(),
             documents_chunk_size: Default::default(),
             max_nb_chunks: Default::default(),
+            experimental_no_edition_2024_for_prefix_post_processing: other
+                .experimental_no_edition_2024_for_prefix_post_processing,
+            experimental_no_edition_2024_for_facet_post_processing: other
+                .experimental_no_edition_2024_for_facet_post_processing,
         })
     }
 }

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

@@ -0,0 +1,84 @@
+use actix_web::web::{self, Data};
+use actix_web::{HttpRequest, HttpResponse};
+use index_scheduler::IndexScheduler;
+use meilisearch_types::error::ResponseError;
+use meilisearch_types::index_uid::IndexUid;
+use meilisearch_types::keys::actions;
+use meilisearch_types::tasks::KindWithContent;
+use tracing::debug;
+use utoipa::OpenApi;
+
+use super::ActionPolicy;
+use crate::analytics::Analytics;
+use crate::extractors::authentication::GuardedData;
+use crate::extractors::sequential_extractor::SeqHandler;
+use crate::routes::SummarizedTaskView;
+
+#[derive(OpenApi)]
+#[openapi(
+    paths(compact),
+    tags(
+        (
+            name = "Compact an index",
+            description = "The /compact route uses compacts the database to reorganize and make it smaller and more efficient.",
+            external_docs(url = "https://www.meilisearch.com/docs/reference/api/compact"),
+        ),
+    ),
+)]
+pub struct CompactApi;
+
+pub fn configure(cfg: &mut web::ServiceConfig) {
+    cfg.service(web::resource("").route(web::post().to(SeqHandler(compact))));
+}
+
+/// Compact an index
+#[utoipa::path(
+    post,
+    path = "{indexUid}/compact",
+    tag = "Compact an index",
+    security(("Bearer" = ["search", "*"])),
+    params(("indexUid" = String, Path, example = "movies", description = "Index Unique Identifier", nullable = false)),
+    responses(
+        (status = ACCEPTED, description = "Task successfully enqueued", body = SummarizedTaskView, content_type = "application/json", example = json!(
+            {
+                "taskUid": 147,
+                "indexUid": null,
+                "status": "enqueued",
+                "type": "documentDeletion",
+                "enqueuedAt": "2024-08-08T17:05:55.791772Z"
+            }
+        )),
+        (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"
+            }
+        )),
+    )
+)]
+pub async fn compact(
+    index_scheduler: GuardedData<ActionPolicy<{ actions::INDEXES_COMPACT }>, Data<IndexScheduler>>,
+    index_uid: web::Path<String>,
+    req: HttpRequest,
+    analytics: web::Data<Analytics>,
+) -> Result<HttpResponse, ResponseError> {
+    let index_uid = IndexUid::try_from(index_uid.into_inner())?;
+
+    analytics.publish(IndexCompacted::default(), &req);
+
+    let task = KindWithContent::IndexCompaction { index_uid: index_uid.to_string() };
+    let task =
+        match tokio::task::spawn_blocking(move || index_scheduler.register(task, None, false))
+            .await?
+        {
+            Ok(task) => task,
+            Err(e) => return Err(e.into()),
+        };
+
+    debug!(returns = ?task, "Compact the {index_uid} index");
+    Ok(HttpResponse::Accepted().json(SummarizedTaskView::from(task)))
+}
+
+crate::empty_analytics!(IndexCompacted, "Index Compacted");

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

@@ -28,6 +28,7 @@ use crate::extractors::sequential_extractor::SeqHandler;
 use crate::routes::is_dry_run;
 use crate::Opt;
 
+pub mod compact;
 pub mod documents;
 mod enterprise_edition;
 pub mod facet_search;
@@ -49,8 +50,9 @@ pub use enterprise_edition::proxy::{PROXY_ORIGIN_REMOTE_HEADER, PROXY_ORIGIN_TAS
         (path = "/", api = facet_search::FacetSearchApi),
         (path = "/", api = similar::SimilarApi),
         (path = "/", api = settings::SettingsApi),
+        (path = "/", api = compact::CompactApi),
     ),
-    paths(list_indexes, create_index, get_index, update_index, delete_index, get_index_stats),
+    paths(list_indexes, create_index, get_index, update_index, delete_index, get_index_stats, compact::compact),
     tags(
         (
             name = "Indexes",
@@ -80,7 +82,8 @@ pub fn configure(cfg: &mut web::ServiceConfig) {
             .service(web::scope("/search").configure(search::configure))
             .service(web::scope("/facet-search").configure(facet_search::configure))
             .service(web::scope("/similar").configure(similar::configure))
-            .service(web::scope("/settings").configure(settings::configure)),
+            .service(web::scope("/settings").configure(settings::configure))
+            .service(web::scope("/compact").configure(compact::configure)),
     );
 }
 

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

@@ -13,6 +13,7 @@ use meilisearch_types::serde_cs::vec::CS;
 use serde_json::Value;
 use tracing::debug;
 use utoipa::{IntoParams, OpenApi};
+use uuid::Uuid;
 
 use crate::analytics::Analytics;
 use crate::error::MeilisearchHttpError;
@@ -21,11 +22,12 @@ use crate::extractors::authentication::GuardedData;
 use crate::extractors::sequential_extractor::SeqHandler;
 use crate::metrics::MEILISEARCH_DEGRADED_SEARCH_REQUESTS;
 use crate::routes::indexes::search_analytics::{SearchAggregator, SearchGET, SearchPOST};
+use crate::routes::parse_include_metadata_header;
 use crate::search::{
     add_search_rules, perform_search, HybridQuery, MatchingStrategy, RankingScoreThreshold,
-    RetrieveVectors, SearchKind, SearchQuery, SearchResult, SemanticRatio, DEFAULT_CROP_LENGTH,
-    DEFAULT_CROP_MARKER, DEFAULT_HIGHLIGHT_POST_TAG, DEFAULT_HIGHLIGHT_PRE_TAG,
-    DEFAULT_SEARCH_LIMIT, DEFAULT_SEARCH_OFFSET, DEFAULT_SEMANTIC_RATIO,
+    RetrieveVectors, SearchKind, SearchParams, SearchQuery, SearchResult, SemanticRatio,
+    DEFAULT_CROP_LENGTH, DEFAULT_CROP_MARKER, DEFAULT_HIGHLIGHT_POST_TAG,
+    DEFAULT_HIGHLIGHT_PRE_TAG, DEFAULT_SEARCH_LIMIT, DEFAULT_SEARCH_OFFSET, DEFAULT_SEMANTIC_RATIO,
 };
 use crate::search_queue::SearchQueue;
 
@@ -325,7 +327,8 @@ pub async fn search_with_url_query(
     req: HttpRequest,
     analytics: web::Data<Analytics>,
 ) -> Result<HttpResponse, ResponseError> {
-    debug!(parameters = ?params, "Search get");
+    let request_uid = Uuid::now_v7();
+    debug!(request_uid = ?request_uid, parameters = ?params, "Search get");
     let index_uid = IndexUid::try_from(index_uid.into_inner())?;
 
     let mut query: SearchQuery = params.into_inner().try_into()?;
@@ -343,14 +346,20 @@ pub async fn search_with_url_query(
         search_kind(&query, index_scheduler.get_ref(), index_uid.to_string(), &index)?;
     let retrieve_vector = RetrieveVectors::new(query.retrieve_vectors);
     let permit = search_queue.try_get_search_permit().await?;
+    let include_metadata = parse_include_metadata_header(&req);
+
     let search_result = tokio::task::spawn_blocking(move || {
         perform_search(
-            index_uid.to_string(),
+            SearchParams {
+                index_uid: index_uid.to_string(),
+                query,
+                search_kind,
+                retrieve_vectors: retrieve_vector,
+                features: index_scheduler.features(),
+                request_uid,
+                include_metadata,
+            },
             &index,
-            query,
-            search_kind,
-            retrieve_vector,
-            index_scheduler.features(),
         )
     })
     .await;
@@ -363,7 +372,7 @@ pub async fn search_with_url_query(
 
     let search_result = search_result?;
 
-    debug!(returns = ?search_result, "Search get");
+    debug!(request_uid = ?request_uid, returns = ?search_result, "Search get");
     Ok(HttpResponse::Ok().json(search_result))
 }
 
@@ -432,9 +441,10 @@ pub async fn search_with_post(
     analytics: web::Data<Analytics>,
 ) -> Result<HttpResponse, ResponseError> {
     let index_uid = IndexUid::try_from(index_uid.into_inner())?;
+    let request_uid = Uuid::now_v7();
 
     let mut query = params.into_inner();
-    debug!(parameters = ?query, "Search post");
+    debug!(request_uid = ?request_uid, parameters = ?query, "Search post");
 
     // Tenant token search_rules.
     if let Some(search_rules) = index_scheduler.filters().get_index_search_rules(&index_uid) {
@@ -449,15 +459,21 @@ pub async fn search_with_post(
         search_kind(&query, index_scheduler.get_ref(), index_uid.to_string(), &index)?;
     let retrieve_vectors = RetrieveVectors::new(query.retrieve_vectors);
 
+    let include_metadata = parse_include_metadata_header(&req);
+
     let permit = search_queue.try_get_search_permit().await?;
     let search_result = tokio::task::spawn_blocking(move || {
         perform_search(
-            index_uid.to_string(),
+            SearchParams {
+                index_uid: index_uid.to_string(),
+                query,
+                search_kind,
+                retrieve_vectors,
+                features: index_scheduler.features(),
+                request_uid,
+                include_metadata,
+            },
             &index,
-            query,
-            search_kind,
-            retrieve_vectors,
-            index_scheduler.features(),
         )
     })
     .await;
@@ -473,7 +489,7 @@ pub async fn search_with_post(
 
     let search_result = search_result?;
 
-    debug!(returns = ?search_result, "Search post");
+    debug!(request_uid = ?request_uid, returns = ?search_result, "Search post");
     Ok(HttpResponse::Ok().json(search_result))
 }
 

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

@@ -234,6 +234,8 @@ impl<Method: AggregateMethod> SearchAggregator<Method> {
             facet_stats: _,
             degraded,
             used_negative_operator,
+            request_uid: _,
+            metadata: _,
         } = result;
 
         self.total_succeeded = self.total_succeeded.saturating_add(1);

crates/meilisearch/src/routes/metrics.rs

@@ -4,6 +4,7 @@ use index_scheduler::{IndexScheduler, Query};
 use meilisearch_auth::AuthController;
 use meilisearch_types::error::ResponseError;
 use meilisearch_types::keys::actions;
+use meilisearch_types::milli::progress::ProgressStepView;
 use meilisearch_types::tasks::Status;
 use prometheus::{Encoder, TextEncoder};
 use time::OffsetDateTime;
@@ -38,6 +39,12 @@ pub fn configure(config: &mut web::ServiceConfig) {
 # HELP meilisearch_db_size_bytes Meilisearch DB Size In Bytes
 # TYPE meilisearch_db_size_bytes gauge
 meilisearch_db_size_bytes 1130496
+# HELP meilisearch_batch_running_progress_trace The currently running progress trace
+# TYPE meilisearch_batch_running_progress_trace gauge
+meilisearch_batch_running_progress_trace{batch_uid="0",step_name="document"} 0.710618582519409
+meilisearch_batch_running_progress_trace{batch_uid="0",step_name="extracting word proximity"} 0.2222222222222222
+meilisearch_batch_running_progress_trace{batch_uid="0",step_name="indexing"} 0.6666666666666666
+meilisearch_batch_running_progress_trace{batch_uid="0",step_name="processing tasks"} 0
 # HELP meilisearch_http_requests_total Meilisearch HTTP requests total
 # TYPE meilisearch_http_requests_total counter
 meilisearch_http_requests_total{method="GET",path="/metrics",status="400"} 1
@@ -61,6 +68,13 @@ meilisearch_http_response_time_seconds_bucket{method="GET",path="/metrics",le="1
 meilisearch_http_response_time_seconds_bucket{method="GET",path="/metrics",le="+Inf"} 0
 meilisearch_http_response_time_seconds_sum{method="GET",path="/metrics"} 0
 meilisearch_http_response_time_seconds_count{method="GET",path="/metrics"} 0
+# HELP meilisearch_last_finished_batches_progress_trace_ms The last few batches progress trace in milliseconds
+# TYPE meilisearch_last_finished_batches_progress_trace_ms gauge
+meilisearch_last_finished_batches_progress_trace_ms{batch_uid="0",step_name="processing tasks"} 19360
+meilisearch_last_finished_batches_progress_trace_ms{batch_uid="0",step_name="processing tasks > computing document changes"} 368
+meilisearch_last_finished_batches_progress_trace_ms{batch_uid="0",step_name="processing tasks > computing document changes > preparing payloads"} 367
+meilisearch_last_finished_batches_progress_trace_ms{batch_uid="0",step_name="processing tasks > computing document changes > preparing payloads > payload"} 367
+meilisearch_last_finished_batches_progress_trace_ms{batch_uid="0",step_name="processing tasks > indexing"} 18970
 # HELP meilisearch_index_count Meilisearch Index Count
 # TYPE meilisearch_index_count gauge
 meilisearch_index_count 1
@@ -148,6 +162,46 @@ pub async fn get_metrics(
         }
     }
 
+    // Fetch and expose the current progressing step
+    crate::metrics::MEILISEARCH_BATCH_RUNNING_PROGRESS_TRACE.reset();
+    let (batches, _total) = index_scheduler.get_batches_from_authorized_indexes(
+        &Query { statuses: Some(vec![Status::Processing]), ..Query::default() },
+        auth_filters,
+    )?;
+    if let Some(batch) = batches.into_iter().next() {
+        let batch_uid = batch.uid.to_string();
+        if let Some(progress) = batch.progress {
+            for ProgressStepView { current_step, finished, total } in progress.steps {
+                crate::metrics::MEILISEARCH_BATCH_RUNNING_PROGRESS_TRACE
+                    .with_label_values(&[batch_uid.as_str(), current_step.as_ref()])
+                    // We return the completion ratio of the current step
+                    .set(finished as f64 / total as f64);
+            }
+        }
+    }
+
+    crate::metrics::MEILISEARCH_LAST_FINISHED_BATCHES_PROGRESS_TRACE_MS.reset();
+    let (batches, _total) = index_scheduler.get_batches_from_authorized_indexes(
+        // Fetch the finished batches...
+        &Query { statuses: Some(vec![Status::Succeeded, Status::Failed]), ..Query::default() },
+        auth_filters,
+    )?;
+    // ...and get the last three batches only.
+    for batch in batches.into_iter().take(3) {
+        let batch_uid = batch.uid.to_string();
+        for (step_name, duration_str) in batch.stats.progress_trace {
+            let Some(duration_str) = duration_str.as_str() else { continue };
+            match humantime::parse_duration(duration_str) {
+                Ok(duration) => {
+                    crate::metrics::MEILISEARCH_LAST_FINISHED_BATCHES_PROGRESS_TRACE_MS
+                        .with_label_values(&[&batch_uid, &step_name])
+                        .set(duration.as_millis() as i64);
+                }
+                Err(e) => tracing::error!("Failed to parse duration: {e}"),
+            }
+        }
+    }
+
     if let Some(last_update) = response.last_update {
         crate::metrics::MEILISEARCH_LAST_UPDATE.set(last_update.unix_timestamp());
     }

crates/meilisearch/src/routes/mod.rs

@@ -45,6 +45,7 @@ use crate::routes::webhooks::{WebhookResults, WebhookSettings, WebhookWithMetada
 use crate::search::{
     FederatedSearch, FederatedSearchResult, Federation, FederationOptions, MergeFacets,
     SearchQueryWithIndex, SearchResultWithIndex, SimilarQuery, SimilarResult,
+    INCLUDE_METADATA_HEADER,
 };
 use crate::search_queue::SearchQueue;
 use crate::Opt;
@@ -184,6 +185,18 @@ pub fn is_dry_run(req: &HttpRequest, opt: &Opt) -> Result<bool, ResponseError> {
         .is_some_and(|s| s.to_lowercase() == "true"))
 }
 
+/// Parse the `Meili-Include-Metadata` header from an HTTP request.
+///
+/// Returns `true` if the header is present and set to "true" or "1" (case-insensitive).
+/// Returns `false` if the header is not present or has any other value.
+pub fn parse_include_metadata_header(req: &HttpRequest) -> bool {
+    req.headers()
+        .get(INCLUDE_METADATA_HEADER)
+        .and_then(|h| h.to_str().ok())
+        .map(|v| matches!(v.to_lowercase().as_str(), "true" | "1"))
+        .unwrap_or(false)
+}
+
 #[derive(Debug, Serialize, Deserialize, ToSchema)]
 #[serde(rename_all = "camelCase")]
 pub struct SummarizedTaskView {

crates/meilisearch/src/routes/multi_search.rs

@@ -9,6 +9,7 @@ use meilisearch_types::keys::actions;
 use serde::Serialize;
 use tracing::debug;
 use utoipa::{OpenApi, ToSchema};
+use uuid::Uuid;
 
 use super::multi_search_analytics::MultiSearchAggregator;
 use crate::analytics::Analytics;
@@ -17,10 +18,11 @@ use crate::extractors::authentication::policies::ActionPolicy;
 use crate::extractors::authentication::{AuthenticationError, GuardedData};
 use crate::extractors::sequential_extractor::SeqHandler;
 use crate::routes::indexes::search::search_kind;
+use crate::routes::parse_include_metadata_header;
 use crate::search::{
     add_search_rules, perform_federated_search, perform_search, FederatedSearch,
-    FederatedSearchResult, RetrieveVectors, SearchQueryWithIndex, SearchResultWithIndex,
-    PROXY_SEARCH_HEADER, PROXY_SEARCH_HEADER_VALUE,
+    FederatedSearchResult, RetrieveVectors, SearchParams, SearchQueryWithIndex,
+    SearchResultWithIndex, PROXY_SEARCH_HEADER, PROXY_SEARCH_HEADER_VALUE,
 };
 use crate::search_queue::SearchQueue;
 
@@ -151,6 +153,7 @@ pub async fn multi_search_with_post(
     // Since we don't want to process half of the search requests and then get a permit refused
     // we're going to get one permit for the whole duration of the multi-search request.
     let permit = search_queue.try_get_search_permit().await?;
+    let request_uid = Uuid::now_v7();
 
     let federated_search = params.into_inner();
 
@@ -186,16 +189,31 @@ pub async fn multi_search_with_post(
         err
     })?;
 
+    let include_metadata = parse_include_metadata_header(&req);
     let response = match federation {
         Some(federation) => {
+            debug!(
+                request_uid = ?request_uid,
+                federation = ?federation,
+                parameters = ?queries,
+                "Federated-search"
+            );
+
             // 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;
+            let search_result = perform_federated_search(
+                &index_scheduler,
+                queries,
+                federation,
+                features,
+                is_proxy,
+                request_uid,
+                include_metadata,
+            )
+            .await;
             permit.drop().await;
 
             if search_result.is_ok() {
@@ -203,6 +221,13 @@ pub async fn multi_search_with_post(
             }
 
             analytics.publish(multi_aggregate, &req);
+
+            debug!(
+                request_uid = ?request_uid,
+                returns = ?search_result,
+                "Federated-search"
+            );
+
             HttpResponse::Ok().json(search_result?)
         }
         None => {
@@ -216,7 +241,12 @@ pub async fn multi_search_with_post(
                     .map(SearchQueryWithIndex::into_index_query_federation)
                     .enumerate()
                 {
-                    debug!(on_index = query_index, parameters = ?query, "Multi-search");
+                    debug!(
+                        request_uid = ?request_uid,
+                        on_index = query_index,
+                        parameters = ?query,
+                        "Multi-search"
+                    );
 
                     if federation_options.is_some() {
                         return Err((
@@ -252,12 +282,16 @@ pub async fn multi_search_with_post(
 
                     let search_result = tokio::task::spawn_blocking(move || {
                         perform_search(
-                            index_uid_str.clone(),
+                            SearchParams {
+                                index_uid: index_uid_str.clone(),
+                                query,
+                                search_kind,
+                                retrieve_vectors: retrieve_vector,
+                                features,
+                                request_uid,
+                                include_metadata,
+                            },
                             &index,
-                            query,
-                            search_kind,
-                            retrieve_vector,
-                            features,
                         )
                     })
                     .await
@@ -286,7 +320,11 @@ pub async fn multi_search_with_post(
                 err
             })?;
 
-            debug!(returns = ?search_results, "Multi-search");
+            debug!(
+                request_uid = ?request_uid,
+                returns = ?search_results,
+                "Multi-search"
+            );
 
             HttpResponse::Ok().json(SearchResults { results: search_results })
         }

crates/meilisearch/src/routes/tasks_test.rs

@@ -226,14 +226,14 @@ mod tests {
         {
             let params = "types=createIndex";
             let err = deserr_query_params::<TaskDeletionOrCancelationQuery>(params).unwrap_err();
-            snapshot!(meili_snap::json_string!(err), @r#"
+            snapshot!(meili_snap::json_string!(err), @r###"
             {
-              "message": "Invalid value in parameter `types`: `createIndex` is not a valid task type. Available types are `documentAdditionOrUpdate`, `documentEdition`, `documentDeletion`, `settingsUpdate`, `indexCreation`, `indexDeletion`, `indexUpdate`, `indexSwap`, `taskCancelation`, `taskDeletion`, `dumpCreation`, `snapshotCreation`, `export`, `upgradeDatabase`.",
+              "message": "Invalid value in parameter `types`: `createIndex` is not a valid task type. Available types are `documentAdditionOrUpdate`, `documentEdition`, `documentDeletion`, `settingsUpdate`, `indexCreation`, `indexDeletion`, `indexUpdate`, `indexSwap`, `taskCancelation`, `taskDeletion`, `dumpCreation`, `snapshotCreation`, `export`, `upgradeDatabase`, `indexCompaction`.",
               "code": "invalid_task_types",
               "type": "invalid_request",
               "link": "https://docs.meilisearch.com/errors#invalid_task_types"
             }
-            "#);
+            "###);
         }
     }
     #[test]

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

@@ -17,11 +17,13 @@ use meilisearch_types::milli::vector::Embedding;
 use meilisearch_types::milli::{self, DocumentId, OrderBy, TimeBudget, DEFAULT_VALUES_PER_FACET};
 use roaring::RoaringBitmap;
 use tokio::task::JoinHandle;
+use uuid::Uuid;
 
 use super::super::ranking_rules::{self, RankingRules};
 use super::super::{
     compute_facet_distribution_stats, prepare_search, AttributesFormat, ComputedFacets, HitMaker,
-    HitsInfo, RetrieveVectors, SearchHit, SearchKind, SearchQuery, SearchQueryWithIndex,
+    HitsInfo, RetrieveVectors, SearchHit, SearchKind, SearchMetadata, SearchQuery,
+    SearchQueryWithIndex,
 };
 use super::proxy::{proxy_search, ProxySearchError, ProxySearchParams};
 use super::types::{
@@ -39,32 +41,58 @@ pub async fn perform_federated_search(
     federation: Federation,
     features: RoFeatures,
     is_proxy: bool,
+    request_uid: Uuid,
+    include_metadata: bool,
 ) -> Result<FederatedSearchResult, ResponseError> {
     if is_proxy {
         features.check_network("Performing a remote federated search")?;
     }
     let before_search = std::time::Instant::now();
-    let deadline = before_search + std::time::Duration::from_secs(9);
+
+    let timeout = std::env::var("MEILI_EXPERIMENTAL_REMOTE_SEARCH_TIMEOUT_SECONDS")
+        .ok()
+        .map(|p| p.parse().unwrap())
+        .unwrap_or(25);
+
+    let deadline = before_search + std::time::Duration::from_secs(timeout);
 
     let required_hit_count = federation.limit + federation.offset;
     let retrieve_vectors = queries.iter().any(|q| q.retrieve_vectors);
 
     let network = index_scheduler.network();
 
+    // Preconstruct metadata keeping the original queries order for later metadata building
+    let precomputed_query_metadata: Option<Vec<_>> = include_metadata.then(|| {
+        queries
+            .iter()
+            .map(|q| {
+                (
+                    q.index_uid.to_string(),
+                    q.federation_options.as_ref().and_then(|o| o.remote.clone()),
+                )
+            })
+            .collect()
+    });
+
     // 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 host and index
     let mut partitioned_queries = PartitionedQueries::new();
+
     for (query_index, federated_query) in queries.into_iter().enumerate() {
         partitioned_queries.partition(federated_query, query_index, &network, features)?
     }
 
     // 2. perform queries, merge and make hits index by index
     // 2.1. start remote queries
-    let remote_search =
-        RemoteSearch::start(partitioned_queries.remote_queries_by_host, &federation, deadline);
+    let remote_search = RemoteSearch::start(
+        partitioned_queries.remote_queries_by_host,
+        &federation,
+        deadline,
+        include_metadata,
+    );
 
     // 2.2. concurrently execute local queries
     let params = SearchByIndexParams {
@@ -106,11 +134,25 @@ pub async fn perform_federated_search(
     let after_waiting_remote_results = std::time::Instant::now();
 
     // 3. merge hits and metadata across indexes and hosts
-    // 3.1. merge metadata
+
+    // 3.1. Build metadata in the same order as the original queries
+    let query_metadata = precomputed_query_metadata.map(|precomputed_query_metadata| {
+        // If a remote is present, set the local remote name
+        let local_remote_name = network.local.clone().filter(|_| partitioned_queries.has_remote);
+
+        build_query_metadata(
+            precomputed_query_metadata,
+            local_remote_name,
+            &remote_results,
+            &results_by_index,
+        )
+    });
+
+    // 3.2. merge federation metadata
     let (estimated_total_hits, degraded, used_negative_operator, facets, max_remote_duration) =
         merge_metadata(&mut results_by_index, &remote_results);
 
-    // 3.2. merge hits
+    // 3.3. merge hits
     let merged_hits: Vec<_> = merge_index_global_results(results_by_index, &mut remote_results)
         .skip(federation.offset)
         .take(federation.limit)
@@ -125,7 +167,7 @@ pub async fn perform_federated_search(
         .map(|hit| hit.hit())
         .collect();
 
-    // 3.3. merge query vectors
+    // 3.4. merge query vectors
     let query_vectors = if retrieve_vectors {
         for remote_results in remote_results.iter_mut() {
             if let Some(remote_vectors) = remote_results.query_vectors.take() {
@@ -144,7 +186,7 @@ pub async fn perform_federated_search(
         None
     };
 
-    // 3.4. merge facets
+    // 3.5. merge facets
     let (facet_distribution, facet_stats, facets_by_index) =
         facet_order.merge(federation.merge_facets, remote_results, facets);
 
@@ -170,6 +212,8 @@ pub async fn perform_federated_search(
         facet_stats,
         facets_by_index,
         remote_errors: partitioned_queries.has_remote.then_some(remote_errors),
+        request_uid: Some(request_uid),
+        metadata: query_metadata,
     })
 }
 
@@ -393,6 +437,7 @@ struct SearchHitByIndex {
 
 struct SearchResultByIndex {
     index: String,
+    primary_key: Option<String>,
     hits: Vec<SearchHitByIndex>,
     estimated_total_hits: usize,
     degraded: bool,
@@ -400,6 +445,61 @@ struct SearchResultByIndex {
     facets: Option<ComputedFacets>,
 }
 
+/// Builds query metadata for federated search results.
+///
+/// This function creates metadata for each query in the same order as the original queries,
+/// combining information from both local and remote search results. It handles the mapping
+/// of primary keys to their respective indexes and remotes to prevent collisions when
+/// multiple remotes have the same index_uid but different primary keys.
+fn build_query_metadata(
+    precomputed_query_metadata: Vec<(String, Option<String>)>,
+    local_remote_name: Option<String>,
+    remote_results: &[FederatedSearchResult],
+    results_by_index: &[SearchResultByIndex],
+) -> Vec<SearchMetadata> {
+    // Create a map of (remote, index_uid) -> primary_key for quick lookup
+    // This prevents collisions when multiple remotes have the same index_uid but different primary keys
+    let mut primary_key_per_index = std::collections::HashMap::new();
+
+    // Build metadata for remote results
+    for remote_result in remote_results {
+        if let Some(remote_metadata) = &remote_result.metadata {
+            for remote_meta in remote_metadata {
+                if let SearchMetadata {
+                    remote: Some(remote_name),
+                    index_uid,
+                    primary_key: Some(primary_key),
+                    ..
+                } = remote_meta
+                {
+                    let key = (Some(remote_name), index_uid);
+                    primary_key_per_index.insert(key, primary_key);
+                }
+            }
+        }
+    }
+
+    // Build metadata for local results
+    for local_meta in results_by_index {
+        if let SearchResultByIndex { index, primary_key: Some(primary_key), .. } = local_meta {
+            let key = (None, index);
+            primary_key_per_index.insert(key, primary_key);
+        }
+    }
+
+    // Build metadata in the same order as the original queries
+    let mut query_metadata = Vec::new();
+    for (index_uid, remote) in precomputed_query_metadata {
+        let primary_key =
+            primary_key_per_index.get(&(remote.as_ref(), &index_uid)).map(|pk| pk.to_string());
+        let query_uid = Uuid::now_v7();
+        // if the remote is not set, use the local remote name
+        let remote = remote.or_else(|| local_remote_name.clone());
+        query_metadata.push(SearchMetadata { query_uid, primary_key, index_uid, remote });
+    }
+    query_metadata
+}
+
 fn merge_metadata(
     results_by_index: &mut Vec<SearchResultByIndex>,
     remote_results: &Vec<FederatedSearchResult>,
@@ -411,6 +511,7 @@ fn merge_metadata(
     let mut max_remote_duration = Duration::ZERO;
     for SearchResultByIndex {
         index,
+        primary_key: _,
         hits: _,
         estimated_total_hits: estimated_total_hits_by_index,
         facets: facets_by_index,
@@ -439,6 +540,8 @@ fn merge_metadata(
         degraded: degraded_for_host,
         used_negative_operator: host_used_negative_operator,
         remote_errors: _,
+        metadata: _,
+        request_uid: _,
     } in remote_results
     {
         let this_remote_duration = Duration::from_millis(*processing_time_ms as u64);
@@ -566,7 +669,12 @@ struct RemoteSearch {
 }
 
 impl RemoteSearch {
-    fn start(queries: RemoteQueriesByHost, federation: &Federation, deadline: Instant) -> Self {
+    fn start(
+        queries: RemoteQueriesByHost,
+        federation: &Federation,
+        deadline: Instant,
+        include_metadata: bool,
+    ) -> Self {
         let mut in_flight_remote_queries = BTreeMap::new();
         let client = reqwest::ClientBuilder::new()
             .connect_timeout(std::time::Duration::from_millis(200))
@@ -586,7 +694,10 @@ impl RemoteSearch {
                     // never merge distant facets
                     proxy_federation.merge_facets = None;
                     let params = params.clone();
-                    async move { proxy_search(&node, queries, proxy_federation, &params).await }
+                    async move {
+                        proxy_search(&node, queries, proxy_federation, &params, include_metadata)
+                            .await
+                    }
                 }),
             );
         }
@@ -630,6 +741,13 @@ impl RemoteSearch {
                             continue 'remote_queries;
                         }
 
+                        // Add remote name to metadata
+                        if let Some(metadata) = res.metadata.as_mut() {
+                            for meta in metadata {
+                                meta.remote = Some(node_name.clone());
+                            }
+                        }
+
                         federation.insert(
                             FEDERATION_REMOTE.to_string(),
                             serde_json::Value::String(node_name.clone()),
@@ -725,6 +843,7 @@ impl SearchByIndex {
             }
         };
         let rtxn = index.read_txn()?;
+        let primary_key = index.primary_key(&rtxn)?.map(|pk| pk.to_string());
         let criteria = index.criteria(&rtxn)?;
         let dictionary = index.dictionary(&rtxn)?;
         let dictionary: Option<Vec<_>> =
@@ -751,6 +870,12 @@ impl SearchByIndex {
             return Err(error);
         }
         let mut results_by_query = Vec::with_capacity(queries.len());
+
+        // all queries for an index share the same budget
+        let time_budget = match cutoff {
+            Some(cutoff) => TimeBudget::new(Duration::from_millis(cutoff)),
+            None => TimeBudget::default(),
+        };
         for QueryByIndex { query, weight, query_index } in queries {
             // use an immediately invoked lambda to capture the result without returning from the function
 
@@ -820,17 +945,13 @@ impl SearchByIndex {
 
                 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,
+                    // clones of `TimeBudget` share the budget rather than restart it
+                    time_budget.clone(),
                     params.features,
                 )?;
 
@@ -977,6 +1098,7 @@ impl SearchByIndex {
             })?;
         self.results_by_index.push(SearchResultByIndex {
             index: index_uid,
+            primary_key,
             hits: merged_result,
             estimated_total_hits,
             degraded,

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

@@ -7,7 +7,7 @@ use serde::de::DeserializeOwned;
 use serde_json::Value;
 
 use super::types::{FederatedSearch, FederatedSearchResult, Federation};
-use crate::search::SearchQueryWithIndex;
+use crate::search::{SearchQueryWithIndex, INCLUDE_METADATA_HEADER};
 
 pub const PROXY_SEARCH_HEADER: &str = "Meili-Proxy-Search";
 pub const PROXY_SEARCH_HEADER_VALUE: &str = "true";
@@ -98,6 +98,7 @@ pub async fn proxy_search(
     queries: Vec<SearchQueryWithIndex>,
     federation: Federation,
     params: &ProxySearchParams,
+    include_metadata: bool,
 ) -> Result<FederatedSearchResult, ProxySearchError> {
     let url = format!("{}/multi-search", node.url);
 
@@ -105,7 +106,12 @@ pub async fn proxy_search(
 
     let search_api_key = node.search_api_key.as_deref();
 
-    let max_deadline = std::time::Instant::now() + std::time::Duration::from_secs(5);
+    let timeout = std::env::var("MEILI_EXPERIMENTAL_REMOTE_SEARCH_TIMEOUT_SECONDS")
+        .ok()
+        .map(|p| p.parse().unwrap())
+        .unwrap_or(25);
+
+    let max_deadline = std::time::Instant::now() + std::time::Duration::from_secs(timeout);
 
     let deadline = if let Some(deadline) = params.deadline {
         std::time::Instant::min(deadline, max_deadline)
@@ -114,7 +120,16 @@ pub async fn proxy_search(
     };
 
     for i in 0..params.try_count {
-        match try_proxy_search(&url, search_api_key, &federated, &params.client, deadline).await {
+        match try_proxy_search(
+            &url,
+            search_api_key,
+            &federated,
+            &params.client,
+            deadline,
+            include_metadata,
+        )
+        .await
+        {
             Ok(response) => return Ok(response),
             Err(retry) => {
                 let duration = retry.into_duration(i)?;
@@ -122,7 +137,7 @@ pub async fn proxy_search(
             }
         }
     }
-    try_proxy_search(&url, search_api_key, &federated, &params.client, deadline)
+    try_proxy_search(&url, search_api_key, &federated, &params.client, deadline, include_metadata)
         .await
         .map_err(Retry::into_error)
 }
@@ -133,6 +148,7 @@ async fn try_proxy_search(
     federated: &FederatedSearch,
     client: &Client,
     deadline: std::time::Instant,
+    include_metadata: bool,
 ) -> Result<FederatedSearchResult, Retry> {
     let timeout = deadline.saturating_duration_since(std::time::Instant::now());
 
@@ -143,6 +159,8 @@ async fn try_proxy_search(
         request
     };
     let request = request.header(PROXY_SEARCH_HEADER, PROXY_SEARCH_HEADER_VALUE);
+    let request =
+        if include_metadata { request.header(INCLUDE_METADATA_HEADER, "true") } else { request };
 
     let response = request.send().await;
     let response = match response {

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

@@ -16,6 +16,9 @@ use meilisearch_types::milli::order_by_map::OrderByMap;
 use meilisearch_types::milli::OrderBy;
 use serde::{Deserialize, Serialize};
 use utoipa::ToSchema;
+use uuid::Uuid;
+
+use crate::search::SearchMetadata;
 
 use super::super::{ComputedFacets, FacetStats, HitsInfo, SearchHit, SearchQueryWithIndex};
 use crate::milli::vector::Embedding;
@@ -131,6 +134,10 @@ pub struct FederatedSearchResult {
     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 request_uid: Option<Uuid>,
+    #[serde(default, skip_serializing_if = "Option::is_none")]
+    pub metadata: Option<Vec<SearchMetadata>>,
 
     #[serde(default, skip_serializing_if = "Option::is_none")]
     pub remote_errors: Option<BTreeMap<String, ResponseError>>,
@@ -156,6 +163,8 @@ impl fmt::Debug for FederatedSearchResult {
             facet_stats,
             facets_by_index,
             remote_errors,
+            request_uid,
+            metadata,
         } = self;
 
         let mut debug = f.debug_struct("SearchResult");
@@ -188,6 +197,12 @@ impl fmt::Debug for FederatedSearchResult {
         if let Some(remote_errors) = remote_errors {
             debug.field("remote_errors", &remote_errors);
         }
+        if let Some(request_uid) = request_uid {
+            debug.field("request_uid", &request_uid);
+        }
+        if let Some(metadata) = metadata {
+            debug.field("metadata", &metadata);
+        }
 
         debug.finish()
     }

crates/meilisearch/src/search/mod.rs

@@ -36,6 +36,7 @@ use serde_json::{json, Value};
 #[cfg(test)]
 mod mod_test;
 use utoipa::ToSchema;
+use uuid::Uuid;
 
 use crate::error::MeilisearchHttpError;
 
@@ -56,6 +57,7 @@ pub const DEFAULT_CROP_MARKER: fn() -> String = || "…".to_string();
 pub const DEFAULT_HIGHLIGHT_PRE_TAG: fn() -> String = || "<em>".to_string();
 pub const DEFAULT_HIGHLIGHT_POST_TAG: fn() -> String = || "</em>".to_string();
 pub const DEFAULT_SEMANTIC_RATIO: fn() -> SemanticRatio = || SemanticRatio(0.5);
+pub const INCLUDE_METADATA_HEADER: &str = "Meili-Include-Metadata";
 
 #[derive(Clone, Default, PartialEq, Deserr, ToSchema)]
 #[deserr(error = DeserrJsonError, rename_all = camelCase, deny_unknown_fields)]
@@ -835,6 +837,18 @@ pub struct SearchHit {
     pub ranking_score_details: Option<serde_json::Map<String, serde_json::Value>>,
 }
 
+#[derive(Serialize, Deserialize, Debug, Clone, PartialEq, ToSchema)]
+#[serde(rename_all = "camelCase")]
+#[schema(rename_all = "camelCase")]
+pub struct SearchMetadata {
+    pub query_uid: Uuid,
+    pub index_uid: String,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub primary_key: Option<String>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub remote: Option<String>,
+}
+
 #[derive(Serialize, Clone, PartialEq, ToSchema)]
 #[serde(rename_all = "camelCase")]
 #[schema(rename_all = "camelCase")]
@@ -851,6 +865,10 @@ pub struct SearchResult {
     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 = "Option::is_none")]
+    pub request_uid: Option<Uuid>,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub metadata: Option<SearchMetadata>,
 
     #[serde(skip_serializing_if = "Option::is_none")]
     pub semantic_hit_count: Option<u32>,
@@ -872,6 +890,8 @@ impl fmt::Debug for SearchResult {
             hits_info,
             facet_distribution,
             facet_stats,
+            request_uid,
+            metadata,
             semantic_hit_count,
             degraded,
             used_negative_operator,
@@ -901,6 +921,12 @@ impl fmt::Debug for SearchResult {
         if let Some(semantic_hit_count) = semantic_hit_count {
             debug.field("semantic_hit_count", &semantic_hit_count);
         }
+        if let Some(request_uid) = request_uid {
+            debug.field("request_uid", &request_uid);
+        }
+        if let Some(metadata) = metadata {
+            debug.field("metadata", &metadata);
+        }
 
         debug.finish()
     }
@@ -1113,15 +1139,28 @@ pub fn prepare_search<'t>(
     Ok((search, is_finite_pagination, max_total_hits, offset))
 }
 
-pub fn perform_search(
-    index_uid: String,
-    index: &Index,
-    query: SearchQuery,
-    search_kind: SearchKind,
-    retrieve_vectors: RetrieveVectors,
-    features: RoFeatures,
-) -> Result<SearchResult, ResponseError> {
+pub struct SearchParams {
+    pub index_uid: String,
+    pub query: SearchQuery,
+    pub search_kind: SearchKind,
+    pub retrieve_vectors: RetrieveVectors,
+    pub features: RoFeatures,
+    pub request_uid: Uuid,
+    pub include_metadata: bool,
+}
+
+pub fn perform_search(params: SearchParams, index: &Index) -> Result<SearchResult, ResponseError> {
+    let SearchParams {
+        index_uid,
+        query,
+        search_kind,
+        retrieve_vectors,
+        features,
+        request_uid,
+        include_metadata,
+    } = params;
     let before_search = Instant::now();
+    let index_uid_for_metadata = index_uid.clone();
     let rtxn = index.read_txn()?;
     let time_budget = match index.search_cutoff(&rtxn)? {
         Some(cutoff) => TimeBudget::new(Duration::from_millis(cutoff)),
@@ -1142,7 +1181,20 @@ pub fn perform_search(
             query_vector,
         },
         semantic_hit_count,
-    ) = search_from_kind(index_uid, search_kind, search)?;
+    ) = search_from_kind(index_uid.clone(), search_kind, search)?;
+
+    let metadata = if include_metadata {
+        let query_uid = Uuid::now_v7();
+        let primary_key = index.primary_key(&rtxn)?.map(|pk| pk.to_string());
+        Some(SearchMetadata {
+            query_uid,
+            index_uid: index_uid_for_metadata,
+            primary_key,
+            remote: None, // Local searches don't have a remote
+        })
+    } else {
+        None
+    };
 
     let SearchQuery {
         q,
@@ -1225,7 +1277,6 @@ pub fn perform_search(
         .transpose()?
         .map(|ComputedFacets { distribution, stats }| (distribution, stats))
         .unzip();
-
     let result = SearchResult {
         hits: documents,
         hits_info,
@@ -1237,6 +1288,8 @@ pub fn perform_search(
         degraded,
         used_negative_operator,
         semantic_hit_count,
+        request_uid: Some(request_uid),
+        metadata,
     };
     Ok(result)
 }

crates/meilisearch/tests/auth/api_keys.rs

@@ -419,14 +419,14 @@ async fn error_add_api_key_invalid_parameters_actions() {
     let (response, code) = server.add_api_key(content).await;
 
     meili_snap::snapshot!(code, @"400 Bad Request");
-    meili_snap::snapshot!(meili_snap::json_string!(response, { ".createdAt" => "[ignored]", ".updatedAt" => "[ignored]" }), @r#"
+    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`, `export`, `network.get`, `network.update`, `chatCompletions`, `chats.*`, `chats.get`, `chats.delete`, `chatsSettings.*`, `chatsSettings.get`, `chatsSettings.update`, `*.get`, `webhooks.get`, `webhooks.update`, `webhooks.delete`, `webhooks.create`, `webhooks.*`",
+      "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`, `export`, `network.get`, `network.update`, `chatCompletions`, `chats.*`, `chats.get`, `chats.delete`, `chatsSettings.*`, `chatsSettings.get`, `chatsSettings.update`, `*.get`, `webhooks.get`, `webhooks.update`, `webhooks.delete`, `webhooks.create`, `webhooks.*`, `indexes.compact`",
       "code": "invalid_api_key_actions",
       "type": "invalid_request",
       "link": "https://docs.meilisearch.com/errors#invalid_api_key_actions"
     }
-    "#);
+    "###);
 }
 
 #[actix_rt::test]

crates/meilisearch/tests/auth/errors.rs

@@ -91,14 +91,14 @@ async fn create_api_key_bad_actions() {
     // can't parse
     let (response, code) = server.add_api_key(json!({ "actions": ["doggo"] })).await;
     snapshot!(code, @"400 Bad Request");
-    snapshot!(json_string!(response), @r#"
+    snapshot!(json_string!(response), @r###"
     {
-      "message": "Unknown value `doggo` 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`, `export`, `network.get`, `network.update`, `chatCompletions`, `chats.*`, `chats.get`, `chats.delete`, `chatsSettings.*`, `chatsSettings.get`, `chatsSettings.update`, `*.get`, `webhooks.get`, `webhooks.update`, `webhooks.delete`, `webhooks.create`, `webhooks.*`",
+      "message": "Unknown value `doggo` 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`, `export`, `network.get`, `network.update`, `chatCompletions`, `chats.*`, `chats.get`, `chats.delete`, `chatsSettings.*`, `chatsSettings.get`, `chatsSettings.update`, `*.get`, `webhooks.get`, `webhooks.update`, `webhooks.delete`, `webhooks.create`, `webhooks.*`, `indexes.compact`",
       "code": "invalid_api_key_actions",
       "type": "invalid_request",
       "link": "https://docs.meilisearch.com/errors#invalid_api_key_actions"
     }
-    "#);
+    "###);
 }
 
 #[actix_rt::test]

crates/meilisearch/tests/batches/errors.rs

@@ -40,14 +40,14 @@ async fn batch_bad_types() {
 
     let (response, code) = server.batches_filter("types=doggo").await;
     snapshot!(code, @"400 Bad Request");
-    snapshot!(json_string!(response), @r#"
+    snapshot!(json_string!(response), @r###"
     {
-      "message": "Invalid value in parameter `types`: `doggo` is not a valid task type. Available types are `documentAdditionOrUpdate`, `documentEdition`, `documentDeletion`, `settingsUpdate`, `indexCreation`, `indexDeletion`, `indexUpdate`, `indexSwap`, `taskCancelation`, `taskDeletion`, `dumpCreation`, `snapshotCreation`, `export`, `upgradeDatabase`.",
+      "message": "Invalid value in parameter `types`: `doggo` is not a valid task type. Available types are `documentAdditionOrUpdate`, `documentEdition`, `documentDeletion`, `settingsUpdate`, `indexCreation`, `indexDeletion`, `indexUpdate`, `indexSwap`, `taskCancelation`, `taskDeletion`, `dumpCreation`, `snapshotCreation`, `export`, `upgradeDatabase`, `indexCompaction`.",
       "code": "invalid_task_types",
       "type": "invalid_request",
       "link": "https://docs.meilisearch.com/errors#invalid_task_types"
     }
-    "#);
+    "###);
 }
 
 #[actix_rt::test]

crates/meilisearch/tests/common/index.rs

@@ -516,6 +516,18 @@ impl<State> Index<'_, State> {
         self.service.post_encoded(url, query, self.encoder).await
     }
 
+    pub async fn search_with_headers(
+        &self,
+        query: Value,
+        headers: Vec<(&str, &str)>,
+    ) -> (Value, StatusCode) {
+        let url = format!("/indexes/{}/search", urlencode(self.uid.as_ref()));
+        let body = serde_json::to_string(&query).unwrap();
+        let mut all_headers = vec![("content-type", "application/json")];
+        all_headers.extend(headers);
+        self.service.post_str(url, body, all_headers).await
+    }
+
     pub async fn search_get(&self, query: &str) -> (Value, StatusCode) {
         let url = format!("/indexes/{}/search{}", urlencode(self.uid.as_ref()), query);
         self.service.get(url).await

crates/meilisearch/tests/common/server.rs

@@ -390,6 +390,17 @@ impl<State> Server<State> {
         self.service.post("/multi-search", queries).await
     }
 
+    pub async fn multi_search_with_headers(
+        &self,
+        queries: Value,
+        headers: Vec<(&str, &str)>,
+    ) -> (Value, StatusCode) {
+        let body = serde_json::to_string(&queries).unwrap();
+        let mut all_headers = vec![("content-type", "application/json")];
+        all_headers.extend(headers);
+        self.service.post_str("/multi-search", body, all_headers).await
+    }
+
     pub async fn list_indexes_raw(&self, parameters: &str) -> (Value, StatusCode) {
         self.service.get(format!("/indexes{parameters}")).await
     }
@@ -490,6 +501,8 @@ pub fn default_settings(dir: impl AsRef<Path>) -> Opt {
             max_indexing_threads: MaxThreads::from_str("2").unwrap(),
             experimental_no_edition_2024_for_settings: false,
             experimental_no_edition_2024_for_dumps: false,
+            experimental_no_edition_2024_for_prefix_post_processing: false,
+            experimental_no_edition_2024_for_facet_post_processing: false,
         },
         experimental_enable_metrics: false,
         ..Parser::parse_from(None as Option<&str>)

crates/meilisearch/tests/documents/add_documents.rs

@@ -1853,7 +1853,7 @@ async fn add_documents_with_geo_field() {
         .await;
     snapshot!(code, @"200 OK");
     // we are expecting docs 4 and 3 first as they have geo
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]" }),
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]", ".requestUid" => "[uuid]" }),
     @r###"
     {
       "hits": [
@@ -1885,7 +1885,8 @@ async fn add_documents_with_geo_field() {
       "processingTimeMs": "[time]",
       "limit": 20,
       "offset": 0,
-      "estimatedTotalHits": 4
+      "estimatedTotalHits": 4,
+      "requestUid": "[uuid]"
     }
     "###);
 }
@@ -1940,7 +1941,7 @@ async fn update_documents_with_geo_field() {
     let (response, code) = index.search_post(json!({"sort": ["_geoPoint(10,0):asc"]})).await;
     snapshot!(code, @"200 OK");
     // we are expecting docs 4 and 3 first as they have geo
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]" }),
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]", ".requestUid" => "[uuid]" }),
     @r###"
     {
       "hits": [
@@ -1972,7 +1973,8 @@ async fn update_documents_with_geo_field() {
       "processingTimeMs": "[time]",
       "limit": 20,
       "offset": 0,
-      "estimatedTotalHits": 4
+      "estimatedTotalHits": 4,
+      "requestUid": "[uuid]"
     }
     "###);
 
@@ -2044,7 +2046,7 @@ async fn update_documents_with_geo_field() {
     let (response, code) = index.search_post(json!({"sort": ["_geoPoint(10,0):asc"]})).await;
     snapshot!(code, @"200 OK");
     // the search response should not have changed: we are expecting docs 4 and 3 first as they have geo
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]" }),
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]", ".requestUid" => "[uuid]" }),
     @r###"
     {
       "hits": [
@@ -2077,7 +2079,8 @@ async fn update_documents_with_geo_field() {
       "processingTimeMs": "[time]",
       "limit": 20,
       "offset": 0,
-      "estimatedTotalHits": 4
+      "estimatedTotalHits": 4,
+      "requestUid": "[uuid]"
     }
     "###);
 }

crates/meilisearch/tests/documents/geojson/mod.rs

@@ -14,17 +14,18 @@ async fn basic_add_settings_and_geojson_documents() {
     server.wait_task(task.uid()).await.succeeded();
 
     let (response, _) = index.search_get("?filter=_geoPolygon([0,0],[0,2],[2,2],[2,0])").await;
-    snapshot!(response,
-    @r#"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }),
+    @r###"
     {
       "hits": [],
       "query": "",
       "processingTimeMs": "[duration]",
       "limit": 20,
       "offset": 0,
-      "estimatedTotalHits": 0
+      "estimatedTotalHits": 0,
+      "requestUid": "[uuid]"
     }
-    "#);
+    "###);
 
     let lille: serde_json::Value = serde_json::from_str(LILLE).unwrap();
     let documents = json!([
@@ -92,8 +93,8 @@ async fn basic_add_settings_and_geojson_documents() {
     "#);
 
     let (response, _code) = index.search_get("?filter=_geoPolygon([0,0],[0,2],[2,2],[2,0])").await;
-    snapshot!(response,
-    @r#"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }),
+    @r###"
     {
       "hits": [
         {
@@ -111,9 +112,10 @@ async fn basic_add_settings_and_geojson_documents() {
       "processingTimeMs": "[duration]",
       "limit": 20,
       "offset": 0,
-      "estimatedTotalHits": 1
+      "estimatedTotalHits": 1,
+      "requestUid": "[uuid]"
     }
-    "#);
+    "###);
 }
 
 #[actix_rt::test]
@@ -174,8 +176,8 @@ async fn basic_add_geojson_documents_and_settings() {
         index.update_settings(json!({"filterableAttributes": ["_geojson"]})).await;
     server.wait_task(task.uid()).await.succeeded();
     let (response, _code) = index.search_get("?filter=_geoPolygon([0,0],[0,2],[2,2],[2,0])").await;
-    snapshot!(response,
-    @r#"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }),
+    @r###"
     {
       "hits": [
         {
@@ -193,9 +195,10 @@ async fn basic_add_geojson_documents_and_settings() {
       "processingTimeMs": "[duration]",
       "limit": 20,
       "offset": 0,
-      "estimatedTotalHits": 1
+      "estimatedTotalHits": 1,
+      "requestUid": "[uuid]"
     }
-    "#);
+    "###);
 }
 
 #[actix_rt::test]
@@ -292,7 +295,7 @@ async fn geo_bounding_box() {
     let (response, code) =
         index.search_get("?filter=_geoBoundingBox([50.53987503447863,21.43443989912143],[43.76393151539099,0.54979129195425])&attributesToRetrieve=name").await;
     snapshot!(code, @"200 OK");
-    snapshot!(response, @r#"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -351,16 +354,17 @@ async fn geo_bounding_box() {
       "processingTimeMs": "[duration]",
       "limit": 20,
       "offset": 0,
-      "estimatedTotalHits": 17
+      "estimatedTotalHits": 17,
+      "requestUid": "[uuid]"
     }
-    "#);
+    "###);
 
     // Between Russia and Alaska
     let (response, code) = index
         .search_get("?filter=_geoBoundingBox([70,-148],[63,152])&attributesToRetrieve=name")
         .await;
     snapshot!(code, @"200 OK");
-    snapshot!(response, @r#"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -377,46 +381,10 @@ async fn geo_bounding_box() {
       "processingTimeMs": "[duration]",
       "limit": 20,
       "offset": 0,
-      "estimatedTotalHits": 3
+      "estimatedTotalHits": 3,
+      "requestUid": "[uuid]"
     }
-    "#);
-}
-
-#[actix_rt::test]
-async fn geo_radius() {
-    let index = shared_index_geojson_documents().await;
-
-    // 200km around Luxembourg
-    let (response, code) = index
-        .search_get("?filter=_geoRadius(49.4369862,6.5576591,200000)&attributesToRetrieve=name")
-        .await;
-    snapshot!(code, @"200 OK");
-    snapshot!(response, @r#"
-    {
-      "hits": [
-        {
-          "name": "Belgium"
-        },
-        {
-          "name": "Germany"
-        },
-        {
-          "name": "France"
-        },
-        {
-          "name": "Luxembourg"
-        },
-        {
-          "name": "Netherlands"
-        }
-      ],
-      "query": "",
-      "processingTimeMs": "[duration]",
-      "limit": 20,
-      "offset": 0,
-      "estimatedTotalHits": 5
-    }
-    "#);
+    "###);
 }
 
 #[actix_rt::test]

crates/meilisearch/tests/search/filters.rs

@@ -742,7 +742,7 @@ async fn vector_filter_all_embedders() {
             "attributesToRetrieve": ["name"]
         }))
         .await;
-    snapshot!(value, @r#"
+    snapshot!(json_string!(value, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -762,9 +762,10 @@ async fn vector_filter_all_embedders() {
       "processingTimeMs": "[duration]",
       "limit": 20,
       "offset": 0,
-      "estimatedTotalHits": 4
+      "estimatedTotalHits": 4,
+      "requestUid": "[uuid]"
     }
-    "#);
+    "###);
 }
 
 #[actix_rt::test]
@@ -839,7 +840,7 @@ async fn vector_filter_specific_embedder() {
             "attributesToRetrieve": ["name"]
         }))
         .await;
-    snapshot!(value, @r#"
+    snapshot!(json_string!(value, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -859,9 +860,10 @@ async fn vector_filter_specific_embedder() {
       "processingTimeMs": "[duration]",
       "limit": 20,
       "offset": 0,
-      "estimatedTotalHits": 4
+      "estimatedTotalHits": 4,
+      "requestUid": "[uuid]"
     }
-    "#);
+    "###);
 }
 
 #[actix_rt::test]
@@ -874,7 +876,7 @@ async fn vector_filter_user_provided() {
             "attributesToRetrieve": ["name"]
         }))
         .await;
-    snapshot!(value, @r#"
+    snapshot!(json_string!(value, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -885,9 +887,10 @@ async fn vector_filter_user_provided() {
       "processingTimeMs": "[duration]",
       "limit": 20,
       "offset": 0,
-      "estimatedTotalHits": 1
+      "estimatedTotalHits": 1,
+      "requestUid": "[uuid]"
     }
-    "#);
+    "###);
 }
 
 #[actix_rt::test]
@@ -900,7 +903,7 @@ async fn vector_filter_specific_fragment() {
             "attributesToRetrieve": ["name"]
         }))
         .await;
-    snapshot!(value, @r#"
+    snapshot!(json_string!(value, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -914,9 +917,10 @@ async fn vector_filter_specific_fragment() {
       "processingTimeMs": "[duration]",
       "limit": 20,
       "offset": 0,
-      "estimatedTotalHits": 2
+      "estimatedTotalHits": 2,
+      "requestUid": "[uuid]"
     }
-    "#);
+    "###);
 
     let (value, _code) = index
         .search_post(json!({
@@ -924,7 +928,7 @@ async fn vector_filter_specific_fragment() {
             "attributesToRetrieve": ["name"]
         }))
         .await;
-    snapshot!(value, @r#"
+    snapshot!(json_string!(value, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -941,9 +945,10 @@ async fn vector_filter_specific_fragment() {
       "processingTimeMs": "[duration]",
       "limit": 20,
       "offset": 0,
-      "estimatedTotalHits": 3
+      "estimatedTotalHits": 3,
+      "requestUid": "[uuid]"
     }
-    "#);
+    "###);
 }
 
 #[actix_rt::test]
@@ -976,16 +981,17 @@ async fn vector_filter_document_template_but_fragments_used() {
             "attributesToRetrieve": ["name"]
         }))
         .await;
-    snapshot!(value, @r#"
+    snapshot!(json_string!(value, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [],
       "query": "",
       "processingTimeMs": "[duration]",
       "limit": 20,
       "offset": 0,
-      "estimatedTotalHits": 0
+      "estimatedTotalHits": 0,
+      "requestUid": "[uuid]"
     }
-    "#);
+    "###);
 }
 
 #[actix_rt::test]
@@ -1023,7 +1029,7 @@ async fn vector_filter_document_template() {
             "attributesToRetrieve": ["name"]
         }))
         .await;
-    snapshot!(value, @r#"
+    snapshot!(json_string!(value, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -1040,9 +1046,10 @@ async fn vector_filter_document_template() {
       "processingTimeMs": "[duration]",
       "limit": 20,
       "offset": 0,
-      "estimatedTotalHits": 3
+      "estimatedTotalHits": 3,
+      "requestUid": "[uuid]"
     }
-    "#);
+    "###);
 }
 
 #[actix_rt::test]
@@ -1075,7 +1082,7 @@ async fn vector_filter_negation() {
             "attributesToRetrieve": ["name"]
         }))
         .await;
-    snapshot!(value, @r#"
+    snapshot!(json_string!(value, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -1092,9 +1099,10 @@ async fn vector_filter_negation() {
       "processingTimeMs": "[duration]",
       "limit": 20,
       "offset": 0,
-      "estimatedTotalHits": 3
+      "estimatedTotalHits": 3,
+      "requestUid": "[uuid]"
     }
-    "#);
+    "###);
 }
 
 #[actix_rt::test]
@@ -1107,7 +1115,7 @@ async fn vector_filter_or_combination() {
             "attributesToRetrieve": ["name"]
         }))
         .await;
-    snapshot!(value, @r#"
+    snapshot!(json_string!(value, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -1124,9 +1132,10 @@ async fn vector_filter_or_combination() {
       "processingTimeMs": "[duration]",
       "limit": 20,
       "offset": 0,
-      "estimatedTotalHits": 3
+      "estimatedTotalHits": 3,
+      "requestUid": "[uuid]"
     }
-    "#);
+    "###);
 }
 
 #[actix_rt::test]
@@ -1139,7 +1148,7 @@ async fn vector_filter_regenerate() {
             "attributesToRetrieve": ["name"]
         }))
         .await;
-    snapshot!(value, @r#"
+    snapshot!(json_string!(value, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -1156,7 +1165,8 @@ async fn vector_filter_regenerate() {
       "processingTimeMs": "[duration]",
       "limit": 20,
       "offset": 0,
-      "estimatedTotalHits": 3
+      "estimatedTotalHits": 3,
+      "requestUid": "[uuid]"
     }
-    "#);
+    "###);
 }

crates/meilisearch/tests/search/geo.rs

@@ -33,7 +33,7 @@ async fn geo_bounding_box_with_string_and_number() {
             }),
             |response, code| {
                 assert_eq!(code, 200, "{response}");
-                snapshot!(json_string!(response, { ".processingTimeMs" => "[time]" }), @r###"
+                snapshot!(json_string!(response, { ".processingTimeMs" => "[time]", ".requestUid" => "[uuid]" }), @r###"
                 {
                   "hits": [
                     {
@@ -63,7 +63,8 @@ async fn geo_bounding_box_with_string_and_number() {
                   "processingTimeMs": "[time]",
                   "limit": 20,
                   "offset": 0,
-                  "estimatedTotalHits": 2
+                  "estimatedTotalHits": 2,
+                  "requestUid": "[uuid]"
                 }
                 "###);
             },
@@ -84,7 +85,7 @@ async fn bug_4640() {
             }),
             |response, code| {
                 assert_eq!(code, 200, "{response}");
-                snapshot!(json_string!(response, { ".processingTimeMs" => "[time]" }), @r###"
+                snapshot!(json_string!(response, { ".processingTimeMs" => "[time]", ".requestUid" => "[uuid]" }), @r###"
                 {
                   "hits": [
                     {
@@ -123,7 +124,8 @@ async fn bug_4640() {
                   "processingTimeMs": "[time]",
                   "limit": 20,
                   "offset": 0,
-                  "estimatedTotalHits": 3
+                  "estimatedTotalHits": 3,
+                  "requestUid": "[uuid]"
                 }
                 "###);
             },
@@ -147,7 +149,7 @@ async fn geo_asc_with_words() {
         &json!({"q": "jean"}),
         |response, code| {
             assert_eq!(code, 200, "{response}");
-            snapshot!(json_string!(response, { ".processingTimeMs" => "[time]" }), @r###"
+            snapshot!(json_string!(response, { ".processingTimeMs" => "[time]", ".requestUid" => "[uuid]" }), @r###"
             {
               "hits": [
                 {
@@ -179,7 +181,8 @@ async fn geo_asc_with_words() {
               "processingTimeMs": "[time]",
               "limit": 20,
               "offset": 0,
-              "estimatedTotalHits": 3
+              "estimatedTotalHits": 3,
+              "requestUid": "[uuid]"
             }
             "###);
         },
@@ -192,7 +195,7 @@ async fn geo_asc_with_words() {
         &json!({"q": "bob"}),
         |response, code| {
             assert_eq!(code, 200, "{response}");
-            snapshot!(json_string!(response, { ".processingTimeMs" => "[time]" }), @r###"
+            snapshot!(json_string!(response, { ".processingTimeMs" => "[time]", ".requestUid" => "[uuid]" }), @r###"
             {
               "hits": [
                 {
@@ -216,7 +219,8 @@ async fn geo_asc_with_words() {
               "processingTimeMs": "[time]",
               "limit": 20,
               "offset": 0,
-              "estimatedTotalHits": 2
+              "estimatedTotalHits": 2,
+              "requestUid": "[uuid]"
             }
             "###);
         },
@@ -229,7 +233,7 @@ async fn geo_asc_with_words() {
         &json!({"q": "intel"}),
         |response, code| {
             assert_eq!(code, 200, "{response}");
-            snapshot!(json_string!(response, { ".processingTimeMs" => "[time]" }), @r###"
+            snapshot!(json_string!(response, { ".processingTimeMs" => "[time]", ".requestUid" => "[uuid]" }), @r###"
             {
               "hits": [
                 {
@@ -245,7 +249,8 @@ async fn geo_asc_with_words() {
               "processingTimeMs": "[time]",
               "limit": 20,
               "offset": 0,
-              "estimatedTotalHits": 1
+              "estimatedTotalHits": 1,
+              "requestUid": "[uuid]"
             }
             "###);
         },
@@ -269,7 +274,7 @@ async fn geo_sort_with_words() {
       &json!({"q": "jean", "sort": ["_geoPoint(0.0, 0.0):asc"]}),
       |response, code| {
           assert_eq!(code, 200, "{response}");
-          snapshot!(json_string!(response, { ".processingTimeMs" => "[time]" }), @r###"
+          snapshot!(json_string!(response, { ".processingTimeMs" => "[time]", ".requestUid" => "[uuid]" }), @r###"
           {
             "hits": [
               {
@@ -304,7 +309,8 @@ async fn geo_sort_with_words() {
             "processingTimeMs": "[time]",
             "limit": 20,
             "offset": 0,
-            "estimatedTotalHits": 3
+            "estimatedTotalHits": 3,
+            "requestUid": "[uuid]"
           }
           "###);
       },

crates/meilisearch/tests/search/hybrid.rs

@@ -1,4 +1,4 @@
-use meili_snap::snapshot;
+use meili_snap::{json_string, snapshot};
 use once_cell::sync::Lazy;
 
 use crate::common::index::Index;
@@ -148,7 +148,7 @@ async fn simple_search() {
         )
         .await;
     snapshot!(code, @"200 OK");
-    snapshot!(response, @r#"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -209,9 +209,10 @@ async fn simple_search() {
       "limit": 20,
       "offset": 0,
       "estimatedTotalHits": 3,
+      "requestUid": "[uuid]",
       "semanticHitCount": 0
     }
-    "#);
+    "###);
     snapshot!(response["semanticHitCount"], @"0");
 
     let (response, code) = index
@@ -220,7 +221,7 @@ async fn simple_search() {
         )
         .await;
     snapshot!(code, @"200 OK");
-    snapshot!(response, @r#"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -284,9 +285,10 @@ async fn simple_search() {
       "limit": 20,
       "offset": 0,
       "estimatedTotalHits": 3,
+      "requestUid": "[uuid]",
       "semanticHitCount": 2
     }
-    "#);
+    "###);
     snapshot!(response["semanticHitCount"], @"2");
 
     let (response, code) = index
@@ -295,7 +297,7 @@ async fn simple_search() {
         )
         .await;
     snapshot!(code, @"200 OK");
-    snapshot!(response, @r#"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -359,9 +361,10 @@ async fn simple_search() {
       "limit": 20,
       "offset": 0,
       "estimatedTotalHits": 3,
+      "requestUid": "[uuid]",
       "semanticHitCount": 3
     }
-    "#);
+    "###);
     snapshot!(response["semanticHitCount"], @"3");
 }
 

crates/meilisearch/tests/search/locales.rs

@@ -104,7 +104,7 @@ async fn simple_search() {
     // english
     index
         .search(json!({"q": "Atta", "attributesToRetrieve": ["id"]}), |response, code| {
-            snapshot!(response, @r###"
+            snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
             {
               "hits": [
                 {
@@ -115,7 +115,8 @@ async fn simple_search() {
               "processingTimeMs": "[duration]",
               "limit": 20,
               "offset": 0,
-              "estimatedTotalHits": 1
+              "estimatedTotalHits": 1,
+              "requestUid": "[uuid]"
             }
             "###);
             snapshot!(code, @"200 OK");
@@ -125,7 +126,7 @@ async fn simple_search() {
     // japanese
     index
         .search(json!({"q": "進撃", "attributesToRetrieve": ["id"]}), |response, code| {
-            snapshot!(response, @r###"
+            snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
             {
               "hits": [
                 {
@@ -136,7 +137,8 @@ async fn simple_search() {
               "processingTimeMs": "[duration]",
               "limit": 20,
               "offset": 0,
-              "estimatedTotalHits": 1
+              "estimatedTotalHits": 1,
+              "requestUid": "[uuid]"
             }
             "###);
             snapshot!(code, @"200 OK");
@@ -147,7 +149,7 @@ async fn simple_search() {
         .search(
             json!({"q": "進撃", "locales": ["jpn"], "attributesToRetrieve": ["id"]}),
             |response, code| {
-                snapshot!(response, @r#"
+                snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
                 {
                   "hits": [
                     {
@@ -158,9 +160,10 @@ async fn simple_search() {
                   "processingTimeMs": "[duration]",
                   "limit": 20,
                   "offset": 0,
-                  "estimatedTotalHits": 1
+                  "estimatedTotalHits": 1,
+                  "requestUid": "[uuid]"
                 }
-                "#);
+                "###);
                 snapshot!(code, @"200 OK");
             },
         )
@@ -169,7 +172,7 @@ async fn simple_search() {
     // chinese
     index
         .search(json!({"q": "进击", "attributesToRetrieve": ["id"]}), |response, code| {
-            snapshot!(response, @r#"
+            snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
             {
               "hits": [
                 {
@@ -180,9 +183,10 @@ async fn simple_search() {
               "processingTimeMs": "[duration]",
               "limit": 20,
               "offset": 0,
-              "estimatedTotalHits": 1
+              "estimatedTotalHits": 1,
+              "requestUid": "[uuid]"
             }
-            "#);
+            "###);
             snapshot!(code, @"200 OK");
         })
         .await;
@@ -222,7 +226,7 @@ async fn force_locales() {
         .search(
             json!({"q": "\"进击的巨人\"", "attributesToRetrieve": ["id"]}),
             |response, code| {
-                snapshot!(response, @r###"
+                snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
                 {
                   "hits": [
                     {
@@ -233,7 +237,8 @@ async fn force_locales() {
                   "processingTimeMs": "[duration]",
                   "limit": 20,
                   "offset": 0,
-                  "estimatedTotalHits": 1
+                  "estimatedTotalHits": 1,
+                  "requestUid": "[uuid]"
                 }
                 "###);
                 snapshot!(code, @"200 OK");
@@ -246,7 +251,7 @@ async fn force_locales() {
         .search(
             json!({"q": "\"进击的巨人\"", "locales": ["jpn"], "attributesToRetrieve": ["id"]}),
             |response, code| {
-                snapshot!(response, @r###"
+                snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
                 {
                   "hits": [
                     {
@@ -257,7 +262,8 @@ async fn force_locales() {
                   "processingTimeMs": "[duration]",
                   "limit": 20,
                   "offset": 0,
-                  "estimatedTotalHits": 1
+                  "estimatedTotalHits": 1,
+                  "requestUid": "[uuid]"
                 }
                 "###);
                 snapshot!(code, @"200 OK");
@@ -300,7 +306,7 @@ async fn force_locales_with_pattern() {
         .search(
             json!({"q": "\"进击的巨人\"", "attributesToRetrieve": ["id"]}),
             |response, code| {
-                snapshot!(response, @r###"
+                snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
                 {
                   "hits": [
                     {
@@ -311,7 +317,8 @@ async fn force_locales_with_pattern() {
                   "processingTimeMs": "[duration]",
                   "limit": 20,
                   "offset": 0,
-                  "estimatedTotalHits": 1
+                  "estimatedTotalHits": 1,
+                  "requestUid": "[uuid]"
                 }
                 "###);
                 snapshot!(code, @"200 OK");
@@ -324,7 +331,7 @@ async fn force_locales_with_pattern() {
         .search(
             json!({"q": "\"进击的巨人\"", "locales": ["jpn"], "attributesToRetrieve": ["id"]}),
             |response, code| {
-                snapshot!(response, @r###"
+                snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
                 {
                   "hits": [
                     {
@@ -335,7 +342,8 @@ async fn force_locales_with_pattern() {
                   "processingTimeMs": "[duration]",
                   "limit": 20,
                   "offset": 0,
-                  "estimatedTotalHits": 1
+                  "estimatedTotalHits": 1,
+                  "requestUid": "[uuid]"
                 }
                 "###);
                 snapshot!(code, @"200 OK");
@@ -376,14 +384,15 @@ async fn force_locales_with_pattern_nested() {
         .search(
             json!({"q": "\"进击的巨人\"", "locales": ["cmn"], "attributesToRetrieve": ["id"]}),
             |response, code| {
-                snapshot!(response, @r###"
+                snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
                 {
                   "hits": [],
                   "query": "\"进击的巨人\"",
                   "processingTimeMs": "[duration]",
                   "limit": 20,
                   "offset": 0,
-                  "estimatedTotalHits": 0
+                  "estimatedTotalHits": 0,
+                  "requestUid": "[uuid]"
                 }
                 "###);
                 snapshot!(code, @"200 OK");
@@ -396,7 +405,7 @@ async fn force_locales_with_pattern_nested() {
         .search(
             json!({"q": "\"进击的巨人\"", "locales": ["jpn"], "attributesToRetrieve": ["id"]}),
             |response, code| {
-                snapshot!(response, @r###"
+                snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
                 {
                   "hits": [
                     {
@@ -407,7 +416,8 @@ async fn force_locales_with_pattern_nested() {
                   "processingTimeMs": "[duration]",
                   "limit": 20,
                   "offset": 0,
-                  "estimatedTotalHits": 1
+                  "estimatedTotalHits": 1,
+                  "requestUid": "[uuid]"
                 }
                 "###);
                 snapshot!(code, @"200 OK");
@@ -451,14 +461,15 @@ async fn force_different_locales_with_pattern() {
         .search(
             json!({"q": "\"进击的巨人\"", "locales": ["cmn"], "attributesToRetrieve": ["id"]}),
             |response, code| {
-                snapshot!(response, @r###"
+                snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
                 {
                   "hits": [],
                   "query": "\"进击的巨人\"",
                   "processingTimeMs": "[duration]",
                   "limit": 20,
                   "offset": 0,
-                  "estimatedTotalHits": 0
+                  "estimatedTotalHits": 0,
+                  "requestUid": "[uuid]"
                 }
                 "###);
                 snapshot!(code, @"200 OK");
@@ -471,7 +482,7 @@ async fn force_different_locales_with_pattern() {
         .search(
             json!({"q": "\"进击的巨人\"", "locales": ["jpn"], "attributesToRetrieve": ["id"]}),
             |response, code| {
-                snapshot!(response, @r###"
+                snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
                 {
                   "hits": [
                     {
@@ -482,7 +493,8 @@ async fn force_different_locales_with_pattern() {
                   "processingTimeMs": "[duration]",
                   "limit": 20,
                   "offset": 0,
-                  "estimatedTotalHits": 1
+                  "estimatedTotalHits": 1,
+                  "requestUid": "[uuid]"
                 }
                 "###);
                 snapshot!(code, @"200 OK");
@@ -529,14 +541,15 @@ async fn auto_infer_locales_at_search_with_attributes_to_search_on() {
         .search(
             json!({"q": "\"进击的巨人\"", "attributesToRetrieve": ["id"]}),
             |response, code| {
-                snapshot!(response, @r###"
+                snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
                 {
                   "hits": [],
                   "query": "\"进击的巨人\"",
                   "processingTimeMs": "[duration]",
                   "limit": 20,
                   "offset": 0,
-                  "estimatedTotalHits": 0
+                  "estimatedTotalHits": 0,
+                  "requestUid": "[uuid]"
                 }
                 "###);
                 snapshot!(code, @"200 OK");
@@ -549,7 +562,7 @@ async fn auto_infer_locales_at_search_with_attributes_to_search_on() {
             .search(
                 json!({"q": "\"进击的巨人\"", "attributesToRetrieve": ["id"], "attributesToSearchOn": ["name_zh", "description_zh"]}),
                 |response, code| {
-                    snapshot!(response, @r###"
+                    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
                     {
                       "hits": [
                         {
@@ -560,7 +573,8 @@ async fn auto_infer_locales_at_search_with_attributes_to_search_on() {
                       "processingTimeMs": "[duration]",
                       "limit": 20,
                       "offset": 0,
-                      "estimatedTotalHits": 1
+                      "estimatedTotalHits": 1,
+                      "requestUid": "[uuid]"
                     }
                     "###);
                     snapshot!(code, @"200 OK");
@@ -602,7 +616,7 @@ async fn auto_infer_locales_at_search() {
         .search(
             json!({"q": "\"进击的巨人\"", "attributesToRetrieve": ["id"]}),
             |response, code| {
-                snapshot!(response, @r###"
+                snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
                 {
                   "hits": [
                     {
@@ -613,7 +627,8 @@ async fn auto_infer_locales_at_search() {
                   "processingTimeMs": "[duration]",
                   "limit": 20,
                   "offset": 0,
-                  "estimatedTotalHits": 1
+                  "estimatedTotalHits": 1,
+                  "requestUid": "[uuid]"
                 }
                 "###);
                 snapshot!(code, @"200 OK");
@@ -625,30 +640,7 @@ async fn auto_infer_locales_at_search() {
         .search(
             json!({"q": "\"进击的巨人\"", "attributesToRetrieve": ["id"]}),
             |response, code| {
-                snapshot!(response, @r###"
-                    {
-                      "hits": [
-                        {
-                          "id": 853
-                        }
-                      ],
-                      "query": "\"进击的巨人\"",
-                      "processingTimeMs": "[duration]",
-                      "limit": 20,
-                      "offset": 0,
-                      "estimatedTotalHits": 1
-                    }
-                    "###);
-                snapshot!(code, @"200 OK");
-            },
-        )
-        .await;
-
-    index
-        .search(
-            json!({"q": "\"进击的巨人\"", "attributesToRetrieve": ["id"]}),
-            |response, code| {
-                snapshot!(response, @r###"
+                snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
                 {
                   "hits": [
                     {
@@ -659,7 +651,32 @@ async fn auto_infer_locales_at_search() {
                   "processingTimeMs": "[duration]",
                   "limit": 20,
                   "offset": 0,
-                  "estimatedTotalHits": 1
+                  "estimatedTotalHits": 1,
+                  "requestUid": "[uuid]"
+                }
+                "###);
+                snapshot!(code, @"200 OK");
+            },
+        )
+        .await;
+
+    index
+        .search(
+            json!({"q": "\"进击的巨人\"", "attributesToRetrieve": ["id"]}),
+            |response, code| {
+                snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
+                {
+                  "hits": [
+                    {
+                      "id": 853
+                    }
+                  ],
+                  "query": "\"进击的巨人\"",
+                  "processingTimeMs": "[duration]",
+                  "limit": 20,
+                  "offset": 0,
+                  "estimatedTotalHits": 1,
+                  "requestUid": "[uuid]"
                 }
                 "###);
                 snapshot!(code, @"200 OK");
@@ -702,14 +719,15 @@ async fn force_different_locales_with_pattern_nested() {
         .search(
             json!({"q": "\"进击的巨人\"", "locales": ["cmn"], "attributesToRetrieve": ["id"]}),
             |response, code| {
-                snapshot!(response, @r###"
+                snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
                 {
                   "hits": [],
                   "query": "\"进击的巨人\"",
                   "processingTimeMs": "[duration]",
                   "limit": 20,
                   "offset": 0,
-                  "estimatedTotalHits": 0
+                  "estimatedTotalHits": 0,
+                  "requestUid": "[uuid]"
                 }
                 "###);
                 snapshot!(code, @"200 OK");
@@ -722,31 +740,7 @@ async fn force_different_locales_with_pattern_nested() {
         .search(
             json!({"q": "\"进击的巨人\"", "locales": ["jpn"], "attributesToRetrieve": ["id"]}),
             |response, code| {
-                snapshot!(response, @r###"
-                    {
-                      "hits": [
-                        {
-                          "id": 852
-                        }
-                      ],
-                      "query": "\"进击的巨人\"",
-                      "processingTimeMs": "[duration]",
-                      "limit": 20,
-                      "offset": 0,
-                      "estimatedTotalHits": 1
-                    }
-                    "###);
-                snapshot!(code, @"200 OK");
-            },
-        )
-        .await;
-
-    // force japanese
-    index
-        .search(
-            json!({"q": "\"进击的巨人\"", "locales": ["ja"], "attributesToRetrieve": ["id"]}),
-            |response, code| {
-                snapshot!(response, @r###"
+                snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
                 {
                   "hits": [
                     {
@@ -757,7 +751,33 @@ async fn force_different_locales_with_pattern_nested() {
                   "processingTimeMs": "[duration]",
                   "limit": 20,
                   "offset": 0,
-                  "estimatedTotalHits": 1
+                  "estimatedTotalHits": 1,
+                  "requestUid": "[uuid]"
+                }
+                "###);
+                snapshot!(code, @"200 OK");
+            },
+        )
+        .await;
+
+    // force japanese
+    index
+        .search(
+            json!({"q": "\"进击的巨人\"", "locales": ["ja"], "attributesToRetrieve": ["id"]}),
+            |response, code| {
+                snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
+                {
+                  "hits": [
+                    {
+                      "id": 852
+                    }
+                  ],
+                  "query": "\"进击的巨人\"",
+                  "processingTimeMs": "[duration]",
+                  "limit": 20,
+                  "offset": 0,
+                  "estimatedTotalHits": 1,
+                  "requestUid": "[uuid]"
                 }
                 "###);
                 snapshot!(code, @"200 OK");
@@ -799,14 +819,15 @@ async fn settings_change() {
         .search(
             json!({"q": "\"进击的巨人\"", "locales": ["cmn"], "attributesToRetrieve": ["id"]}),
             |response, code| {
-                snapshot!(response, @r###"
+                snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
                 {
                   "hits": [],
                   "query": "\"进击的巨人\"",
                   "processingTimeMs": "[duration]",
                   "limit": 20,
                   "offset": 0,
-                  "estimatedTotalHits": 0
+                  "estimatedTotalHits": 0,
+                  "requestUid": "[uuid]"
                 }
                 "###);
                 snapshot!(code, @"200 OK");
@@ -819,14 +840,15 @@ async fn settings_change() {
         .search(
             json!({"q": "\"进击的巨人\"", "locales": ["jpn"], "attributesToRetrieve": ["id"]}),
             |response, code| {
-                snapshot!(response, @r###"
+                snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
                 {
                   "hits": [],
                   "query": "\"进击的巨人\"",
                   "processingTimeMs": "[duration]",
                   "limit": 20,
                   "offset": 0,
-                  "estimatedTotalHits": 0
+                  "estimatedTotalHits": 0,
+                  "requestUid": "[uuid]"
                 }
                 "###);
                 snapshot!(code, @"200 OK");
@@ -862,14 +884,15 @@ async fn settings_change() {
         .search(
             json!({"q": "\"进击的巨人\"", "locales": ["cmn"], "attributesToRetrieve": ["id"]}),
             |response, code| {
-                snapshot!(response, @r###"
+                snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
                 {
                   "hits": [],
                   "query": "\"进击的巨人\"",
                   "processingTimeMs": "[duration]",
                   "limit": 20,
                   "offset": 0,
-                  "estimatedTotalHits": 0
+                  "estimatedTotalHits": 0,
+                  "requestUid": "[uuid]"
                 }
                 "###);
                 snapshot!(code, @"200 OK");
@@ -882,14 +905,15 @@ async fn settings_change() {
         .search(
             json!({"q": "\"进击的巨人\"", "locales": ["jpn"], "attributesToRetrieve": ["id"]}),
             |response, code| {
-                snapshot!(response, @r###"
+                snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
                 {
                   "hits": [],
                   "query": "\"进击的巨人\"",
                   "processingTimeMs": "[duration]",
                   "limit": 20,
                   "offset": 0,
-                  "estimatedTotalHits": 0
+                  "estimatedTotalHits": 0,
+                  "requestUid": "[uuid]"
                 }
                 "###);
                 snapshot!(code, @"200 OK");
@@ -1164,7 +1188,7 @@ async fn swedish_search() {
     // infer swedish
     index
         .search(json!({"q": "trä", "attributesToRetrieve": ["product"]}), |response, code| {
-            snapshot!(response, @r###"
+            snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
             {
               "hits": [
                 {
@@ -1178,7 +1202,8 @@ async fn swedish_search() {
               "processingTimeMs": "[duration]",
               "limit": 20,
               "offset": 0,
-              "estimatedTotalHits": 2
+              "estimatedTotalHits": 2,
+              "requestUid": "[uuid]"
             }
             "###);
             snapshot!(code, @"200 OK");
@@ -1187,7 +1212,7 @@ async fn swedish_search() {
 
     index
         .search(json!({"q": "tra", "attributesToRetrieve": ["product"]}), |response, code| {
-            snapshot!(response, @r###"
+            snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
             {
               "hits": [
                 {
@@ -1201,7 +1226,8 @@ async fn swedish_search() {
               "processingTimeMs": "[duration]",
               "limit": 20,
               "offset": 0,
-              "estimatedTotalHits": 2
+              "estimatedTotalHits": 2,
+              "requestUid": "[uuid]"
             }
             "###);
             snapshot!(code, @"200 OK");
@@ -1213,7 +1239,7 @@ async fn swedish_search() {
         .search(
             json!({"q": "trä", "locales": ["swe"], "attributesToRetrieve": ["product"]}),
             |response, code| {
-                snapshot!(response, @r###"
+                snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
                 {
                   "hits": [
                     {
@@ -1227,7 +1253,8 @@ async fn swedish_search() {
                   "processingTimeMs": "[duration]",
                   "limit": 20,
                   "offset": 0,
-                  "estimatedTotalHits": 2
+                  "estimatedTotalHits": 2,
+                  "requestUid": "[uuid]"
                 }
                 "###);
                 snapshot!(code, @"200 OK");
@@ -1238,7 +1265,7 @@ async fn swedish_search() {
         .search(
             json!({"q": "tra", "locales": ["swe"], "attributesToRetrieve": ["product"]}),
             |response, code| {
-                snapshot!(response, @r###"
+                snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
                 {
                   "hits": [
                     {
@@ -1252,7 +1279,8 @@ async fn swedish_search() {
                   "processingTimeMs": "[duration]",
                   "limit": 20,
                   "offset": 0,
-                  "estimatedTotalHits": 2
+                  "estimatedTotalHits": 2,
+                  "requestUid": "[uuid]"
                 }
                 "###);
                 snapshot!(code, @"200 OK");
@@ -1287,20 +1315,21 @@ async fn german_search() {
         .search(
             json!({"q": "kulturalität", "attributesToRetrieve": ["product"]}),
             |response, code| {
-                snapshot!(response, @r###"
-            {
-              "hits": [
+                snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
                 {
-                  "product": "Interkulturalität"
+                  "hits": [
+                    {
+                      "product": "Interkulturalität"
+                    }
+                  ],
+                  "query": "kulturalität",
+                  "processingTimeMs": "[duration]",
+                  "limit": 20,
+                  "offset": 0,
+                  "estimatedTotalHits": 1,
+                  "requestUid": "[uuid]"
                 }
-              ],
-              "query": "kulturalität",
-              "processingTimeMs": "[duration]",
-              "limit": 20,
-              "offset": 0,
-              "estimatedTotalHits": 1
-            }
-            "###);
+                "###);
                 snapshot!(code, @"200 OK");
             },
         )
@@ -1310,7 +1339,7 @@ async fn german_search() {
         .search(
             json!({"q": "organisation", "attributesToRetrieve": ["product"]}),
             |response, code| {
-                snapshot!(response, @r###"
+                snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
             {
               "hits": [
                 {
@@ -1321,7 +1350,8 @@ async fn german_search() {
               "processingTimeMs": "[duration]",
               "limit": 20,
               "offset": 0,
-              "estimatedTotalHits": 1
+              "estimatedTotalHits": 1,
+              "requestUid": "[uuid]"
             }
             "###);
                 snapshot!(code, @"200 OK");

crates/meilisearch/tests/search/metadata.rs

@@ -0,0 +1,387 @@
+use meili_snap::{json_string, snapshot};
+
+use crate::common::{shared_index_with_documents, Server, DOCUMENTS};
+use crate::json;
+
+#[actix_rt::test]
+async fn search_without_metadata_header() {
+    let index = shared_index_with_documents().await;
+
+    // Test that metadata is not included by default
+    index
+        .search(json!({"q": "glass"}), |response, code| {
+            snapshot!(code, @"200 OK");
+            snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
+            {
+              "hits": [
+                {
+                  "title": "Gläss",
+                  "id": "450465",
+                  "color": [
+                    "blue",
+                    "red"
+                  ]
+                }
+              ],
+              "query": "glass",
+              "processingTimeMs": "[duration]",
+              "limit": 20,
+              "offset": 0,
+              "estimatedTotalHits": 1,
+              "requestUid": "[uuid]"
+            }
+            "###);
+        })
+        .await;
+}
+
+#[actix_rt::test]
+async fn search_with_metadata_header() {
+    let server = Server::new_shared();
+    let index = server.unique_index();
+
+    let documents = DOCUMENTS.clone();
+    let (task, _code) = index.add_documents(documents, None).await;
+    server.wait_task(task.uid()).await.succeeded();
+
+    // Test with Meili-Include-Metadata header
+    let (response, code) = index
+        .search_with_headers(json!({"q": "glass"}), vec![("Meili-Include-Metadata", "true")])
+        .await;
+
+    snapshot!(code, @"200 OK");
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]", ".metadata.queryUid" => "[uuid]" }), @r###"
+    {
+      "hits": [
+        {
+          "title": "Gläss",
+          "id": "450465",
+          "color": [
+            "blue",
+            "red"
+          ]
+        }
+      ],
+      "query": "glass",
+      "processingTimeMs": "[duration]",
+      "limit": 20,
+      "offset": 0,
+      "estimatedTotalHits": 1,
+      "requestUid": "[uuid]",
+      "metadata": {
+        "queryUid": "[uuid]",
+        "indexUid": "[uuid]",
+        "primaryKey": "id"
+      }
+    }
+    "###);
+}
+
+#[actix_rt::test]
+async fn search_with_metadata_header_and_primary_key() {
+    let server = Server::new_shared();
+    let index = server.unique_index();
+
+    let documents = DOCUMENTS.clone();
+    let (task, _code) = index.add_documents(documents, Some("id")).await;
+    server.wait_task(task.uid()).await.succeeded();
+
+    // Test with Meili-Include-Metadata header
+    let (response, code) = index
+        .search_with_headers(json!({"q": "glass"}), vec![("Meili-Include-Metadata", "true")])
+        .await;
+
+    snapshot!(code, @"200 OK");
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]", ".metadata.queryUid" => "[uuid]" }), @r###"
+    {
+      "hits": [
+        {
+          "id": "450465",
+          "title": "Gläss",
+          "color": [
+            "blue",
+            "red"
+          ]
+        }
+      ],
+      "query": "glass",
+      "processingTimeMs": "[duration]",
+      "limit": 20,
+      "offset": 0,
+      "estimatedTotalHits": 1,
+      "requestUid": "[uuid]",
+      "metadata": {
+        "queryUid": "[uuid]",
+        "indexUid": "[uuid]",
+        "primaryKey": "id"
+      }
+    }
+    "###);
+}
+
+#[actix_rt::test]
+async fn multi_search_without_metadata_header() {
+    let server = Server::new_shared();
+    let index = server.unique_index();
+
+    let documents = DOCUMENTS.clone();
+    let (task, _code) = index.add_documents(documents, None).await;
+    server.wait_task(task.uid()).await.succeeded();
+
+    // Test multi-search without metadata header
+    let (response, code) = server
+        .multi_search(json!({
+            "queries": [
+                {"indexUid": index.uid, "q": "glass"},
+                {"indexUid": index.uid, "q": "dragon"}
+            ]
+        }))
+        .await;
+
+    snapshot!(code, @"200 OK");
+    snapshot!(json_string!(response, { ".results[0].processingTimeMs" => "[duration]", ".results[0].requestUid" => "[uuid]", ".results[1].processingTimeMs" => "[duration]", ".results[1].requestUid" => "[uuid]" }), @r###"
+    {
+      "results": [
+        {
+          "indexUid": "[uuid]",
+          "hits": [
+            {
+              "title": "Gläss",
+              "id": "450465",
+              "color": [
+                "blue",
+                "red"
+              ]
+            }
+          ],
+          "query": "glass",
+          "processingTimeMs": "[duration]",
+          "limit": 20,
+          "offset": 0,
+          "estimatedTotalHits": 1,
+          "requestUid": "[uuid]"
+        },
+        {
+          "indexUid": "[uuid]",
+          "hits": [
+            {
+              "title": "How to Train Your Dragon: The Hidden World",
+              "id": "166428",
+              "color": [
+                "green",
+                "red"
+              ]
+            }
+          ],
+          "query": "dragon",
+          "processingTimeMs": "[duration]",
+          "limit": 20,
+          "offset": 0,
+          "estimatedTotalHits": 1,
+          "requestUid": "[uuid]"
+        }
+      ]
+    }
+    "###);
+}
+
+#[actix_rt::test]
+async fn multi_search_with_metadata_header() {
+    let server = Server::new_shared();
+    let index = server.unique_index();
+
+    let documents = DOCUMENTS.clone();
+    let (task, _code) = index.add_documents(documents, Some("id")).await;
+    server.wait_task(task.uid()).await.succeeded();
+
+    // Test multi-search with metadata header
+    let (response, code) = server
+        .multi_search_with_headers(
+            json!({
+                "queries": [
+                    {"indexUid": index.uid, "q": "glass"},
+                    {"indexUid": index.uid, "q": "dragon"}
+                ]
+            }),
+            vec![("Meili-Include-Metadata", "true")],
+        )
+        .await;
+
+    snapshot!(code, @"200 OK");
+    snapshot!(json_string!(response, { ".results[0].processingTimeMs" => "[duration]", ".results[0].requestUid" => "[uuid]", ".results[0].metadata.queryUid" => "[uuid]", ".results[1].processingTimeMs" => "[duration]", ".results[1].requestUid" => "[uuid]", ".results[1].metadata.queryUid" => "[uuid]" }), @r###"
+    {
+      "results": [
+        {
+          "indexUid": "[uuid]",
+          "hits": [
+            {
+              "id": "450465",
+              "title": "Gläss",
+              "color": [
+                "blue",
+                "red"
+              ]
+            }
+          ],
+          "query": "glass",
+          "processingTimeMs": "[duration]",
+          "limit": 20,
+          "offset": 0,
+          "estimatedTotalHits": 1,
+          "requestUid": "[uuid]",
+          "metadata": {
+            "queryUid": "[uuid]",
+            "indexUid": "[uuid]",
+            "primaryKey": "id"
+          }
+        },
+        {
+          "indexUid": "[uuid]",
+          "hits": [
+            {
+              "id": "166428",
+              "title": "How to Train Your Dragon: The Hidden World",
+              "color": [
+                "green",
+                "red"
+              ]
+            }
+          ],
+          "query": "dragon",
+          "processingTimeMs": "[duration]",
+          "limit": 20,
+          "offset": 0,
+          "estimatedTotalHits": 1,
+          "requestUid": "[uuid]",
+          "metadata": {
+            "queryUid": "[uuid]",
+            "indexUid": "[uuid]",
+            "primaryKey": "id"
+          }
+        }
+      ]
+    }
+    "###);
+}
+
+#[actix_rt::test]
+async fn search_metadata_header_false_value() {
+    let server = Server::new_shared();
+    let index = server.unique_index();
+
+    let documents = DOCUMENTS.clone();
+    let (task, _code) = index.add_documents(documents, None).await;
+    server.wait_task(task.uid()).await.succeeded();
+
+    // Test with header set to false
+    let (response, code) = index
+        .search_with_headers(json!({"q": "glass"}), vec![("Meili-Include-Metadata", "false")])
+        .await;
+
+    snapshot!(code, @"200 OK");
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
+    {
+      "hits": [
+        {
+          "title": "Gläss",
+          "id": "450465",
+          "color": [
+            "blue",
+            "red"
+          ]
+        }
+      ],
+      "query": "glass",
+      "processingTimeMs": "[duration]",
+      "limit": 20,
+      "offset": 0,
+      "estimatedTotalHits": 1,
+      "requestUid": "[uuid]"
+    }
+    "###);
+}
+
+#[actix_rt::test]
+async fn search_metadata_uuid_format() {
+    let server = Server::new_shared();
+    let index = server.unique_index();
+
+    let documents = DOCUMENTS.clone();
+    let (task, _code) = index.add_documents(documents, None).await;
+    server.wait_task(task.uid()).await.succeeded();
+
+    let (response, code) = index
+        .search_with_headers(json!({"q": "glass"}), vec![("Meili-Include-Metadata", "true")])
+        .await;
+
+    snapshot!(code, @"200 OK");
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]", ".metadata.queryUid" => "[uuid]" }), @r###"
+    {
+      "hits": [
+        {
+          "title": "Gläss",
+          "id": "450465",
+          "color": [
+            "blue",
+            "red"
+          ]
+        }
+      ],
+      "query": "glass",
+      "processingTimeMs": "[duration]",
+      "limit": 20,
+      "offset": 0,
+      "estimatedTotalHits": 1,
+      "requestUid": "[uuid]",
+      "metadata": {
+        "queryUid": "[uuid]",
+        "indexUid": "[uuid]",
+        "primaryKey": "id"
+      }
+    }
+    "###);
+}
+
+#[actix_rt::test]
+async fn search_metadata_consistency_across_requests() {
+    let server = Server::new_shared();
+    let index = server.unique_index();
+
+    let documents = DOCUMENTS.clone();
+    let (task, _code) = index.add_documents(documents, Some("id")).await;
+    server.wait_task(task.uid()).await.succeeded();
+
+    // Make multiple requests and check that metadata is consistent
+    for _i in 0..3 {
+        let (response, code) = index
+            .search_with_headers(json!({"q": "glass"}), vec![("Meili-Include-Metadata", "true")])
+            .await;
+
+        snapshot!(code, @"200 OK");
+        snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]", ".metadata.queryUid" => "[uuid]" }), @r###"
+        {
+          "hits": [
+            {
+              "id": "450465",
+              "title": "Gläss",
+              "color": [
+                "blue",
+                "red"
+              ]
+            }
+          ],
+          "query": "glass",
+          "processingTimeMs": "[duration]",
+          "limit": 20,
+          "offset": 0,
+          "estimatedTotalHits": 1,
+          "requestUid": "[uuid]",
+          "metadata": {
+            "queryUid": "[uuid]",
+            "indexUid": "[uuid]",
+            "primaryKey": "id"
+          }
+        }
+        "###);
+    }
+}

crates/meilisearch/tests/search/mod.rs

@@ -11,6 +11,7 @@ mod hybrid;
 #[cfg(not(feature = "chinese-pinyin"))]
 mod locales;
 mod matching_strategy;
+mod metadata;
 mod multi;
 mod pagination;
 mod restrict_searchable;
@@ -1044,7 +1045,7 @@ async fn test_degraded_score_details() {
             }),
             |response, code| {
                 snapshot!(code, @"200 OK");
-                snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]" }), @r###"
+                snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
                 {
                   "hits": [
                     {
@@ -1103,7 +1104,8 @@ async fn test_degraded_score_details() {
                   "processingTimeMs": "[duration]",
                   "limit": 20,
                   "offset": 0,
-                  "estimatedTotalHits": 3
+                  "estimatedTotalHits": 3,
+                  "requestUid": "[uuid]"
                 }
                 "###);
             },
@@ -2127,3 +2129,102 @@ async fn simple_search_changing_unrelated_settings() {
         })
         .await;
 }
+
+#[actix_rt::test]
+async fn ranking_score_bug_with_sort() {
+    let server = Server::new_shared();
+    let index = server.unique_index();
+
+    // Create documents with a "created" field for sorting
+    let documents = json!([
+        {
+            "id": "1",
+            "title": "Coffee Mug",
+            "created": "2023-01-01T00:00:00Z"
+        },
+        {
+            "id": "2",
+            "title": "Water Bottle",
+            "created": "2023-01-02T00:00:00Z"
+        },
+        {
+            "id": "3",
+            "title": "Tumbler Cup",
+            "created": "2023-01-03T00:00:00Z"
+        },
+        {
+            "id": "4",
+            "title": "Stainless Steel Tumbler",
+            "created": "2023-01-04T00:00:00Z"
+        }
+    ]);
+
+    // Add documents
+    let (task, code) = index.add_documents(documents, None).await;
+    assert_eq!(code, 202, "{task}");
+    server.wait_task(task.uid()).await.succeeded();
+
+    // Configure sortable attributes
+    let (task, code) = index
+        .update_settings(json!({
+            "sortableAttributes": ["created"]
+        }))
+        .await;
+    assert_eq!(code, 202, "{task}");
+    server.wait_task(task.uid()).await.succeeded();
+
+    // Test 1: Search without sort - should have proper ranking scores
+    index
+        .search(
+            json!({
+                "q": "tumbler",
+                "showRankingScore": true,
+                "rankingScoreThreshold": 0.0,
+                "attributesToRetrieve": ["title"]
+            }),
+            |response, code| {
+                assert_eq!(code, 200, "{response}");
+                snapshot!(json_string!(response["hits"]), @r###"
+                [
+                  {
+                    "title": "Tumbler Cup",
+                    "_rankingScore": 0.9848484848484848
+                  },
+                  {
+                    "title": "Stainless Steel Tumbler",
+                    "_rankingScore": 0.8787878787878788
+                  }
+                ]
+                "###);
+            },
+        )
+        .await;
+
+    // Test 2: Search with sort - this is where the bug occurs
+    index
+        .search(
+            json!({
+                "q": "tumbler",
+                "showRankingScore": true,
+                "rankingScoreThreshold": 0.0,
+                "sort": ["created:desc"],
+                "attributesToRetrieve": ["title"]
+            }),
+            |response, code| {
+                assert_eq!(code, 200, "{response}");
+                snapshot!(json_string!(response["hits"]), @r###"
+                [
+                  {
+                    "title": "Tumbler Cup",
+                    "_rankingScore": 0.9848484848484848
+                  },
+                  {
+                    "title": "Stainless Steel Tumbler",
+                    "_rankingScore": 0.8787878787878788
+                  }
+                ]
+                "###);
+            },
+        )
+        .await;
+}

crates/meilisearch/tests/search/multi/mod.rs

@@ -93,13 +93,14 @@ async fn federation_empty_list() {
 
     let (response, code) = server.multi_search(json!({"federation": {}, "queries": []})).await;
     snapshot!(code, @"200 OK");
-    snapshot!(response, @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [],
       "processingTimeMs": "[duration]",
       "limit": 20,
       "offset": 0,
-      "estimatedTotalHits": 0
+      "estimatedTotalHits": 0,
+      "requestUid": "[uuid]"
     }
     "###);
 }
@@ -164,7 +165,7 @@ async fn simple_search_single_index() {
         ]}))
         .await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response["results"], { ".**.processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]" }), @r###"
+    snapshot!(json_string!(response["results"], { ".**.processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]", ".**.requestUid" => "[uuid]" }), @r###"
     [
       {
         "indexUid": "SHARED_DOCUMENTS",
@@ -182,7 +183,8 @@ async fn simple_search_single_index() {
         "processingTimeMs": "[duration]",
         "limit": 20,
         "offset": 0,
-        "estimatedTotalHits": 1
+        "estimatedTotalHits": 1,
+        "requestUid": "[uuid]"
       },
       {
         "indexUid": "SHARED_DOCUMENTS",
@@ -200,7 +202,8 @@ async fn simple_search_single_index() {
         "processingTimeMs": "[duration]",
         "limit": 20,
         "offset": 0,
-        "estimatedTotalHits": 1
+        "estimatedTotalHits": 1,
+        "requestUid": "[uuid]"
       }
     ]
     "###);
@@ -217,7 +220,7 @@ async fn federation_single_search_single_index() {
         ]}))
         .await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]", ".**.requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -237,7 +240,8 @@ async fn federation_single_search_single_index() {
       "processingTimeMs": "[duration]",
       "limit": 20,
       "offset": 0,
-      "estimatedTotalHits": 1
+      "estimatedTotalHits": 1,
+      "requestUid": "[uuid]"
     }
     "###);
 }
@@ -256,7 +260,7 @@ async fn federation_multiple_search_single_index() {
         ]}))
         .await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]", ".**.requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -308,7 +312,8 @@ async fn federation_multiple_search_single_index() {
       "processingTimeMs": "[duration]",
       "limit": 20,
       "offset": 0,
-      "estimatedTotalHits": 5
+      "estimatedTotalHits": 5,
+      "requestUid": "[uuid]"
     }
     "###);
 }
@@ -325,7 +330,7 @@ async fn federation_two_search_single_index() {
         ]}))
         .await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]", ".**.requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -358,7 +363,8 @@ async fn federation_two_search_single_index() {
       "processingTimeMs": "[duration]",
       "limit": 20,
       "offset": 0,
-      "estimatedTotalHits": 2
+      "estimatedTotalHits": 2,
+      "requestUid": "[uuid]"
     }
     "###);
 }
@@ -457,7 +463,7 @@ async fn simple_search_two_indexes() {
         ]}))
         .await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response["results"], { ".**.processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]" }), @r###"
+    snapshot!(json_string!(response["results"], { ".**.processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]", ".**.requestUid" => "[uuid]" }), @r###"
     [
       {
         "indexUid": "SHARED_DOCUMENTS",
@@ -475,7 +481,8 @@ async fn simple_search_two_indexes() {
         "processingTimeMs": "[duration]",
         "limit": 20,
         "offset": 0,
-        "estimatedTotalHits": 1
+        "estimatedTotalHits": 1,
+        "requestUid": "[uuid]"
       },
       {
         "indexUid": "SHARED_NESTED_DOCUMENTS",
@@ -516,7 +523,8 @@ async fn simple_search_two_indexes() {
         "processingTimeMs": "[duration]",
         "limit": 20,
         "offset": 0,
-        "estimatedTotalHits": 2
+        "estimatedTotalHits": 2,
+        "requestUid": "[uuid]"
       }
     ]
     "###);
@@ -535,7 +543,7 @@ async fn federation_two_search_two_indexes() {
         ]}))
         .await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]", ".**.requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -596,7 +604,8 @@ async fn federation_two_search_two_indexes() {
       "processingTimeMs": "[duration]",
       "limit": 20,
       "offset": 0,
-      "estimatedTotalHits": 3
+      "estimatedTotalHits": 3,
+      "requestUid": "[uuid]"
     }
     "###);
 }
@@ -626,7 +635,7 @@ async fn federation_multiple_search_multiple_indexes() {
         ]}))
         .await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]", ".**.requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -795,7 +804,8 @@ async fn federation_multiple_search_multiple_indexes() {
       "processingTimeMs": "[duration]",
       "limit": 20,
       "offset": 0,
-      "estimatedTotalHits": 12
+      "estimatedTotalHits": 12,
+      "requestUid": "[uuid]"
     }
     "###);
 }
@@ -1101,7 +1111,7 @@ async fn federation_filter() {
         ]}))
         .await;
     snapshot!(code, @"200 OK");
-    snapshot!(response, @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -1140,7 +1150,8 @@ async fn federation_filter() {
       "processingTimeMs": "[duration]",
       "limit": 20,
       "offset": 0,
-      "estimatedTotalHits": 3
+      "estimatedTotalHits": 3,
+      "requestUid": "[uuid]"
     }
     "###);
 }
@@ -1177,7 +1188,7 @@ async fn federation_sort_same_indexes_same_criterion_same_direction() {
         ]}))
         .await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -1266,7 +1277,8 @@ async fn federation_sort_same_indexes_same_criterion_same_direction() {
       "processingTimeMs": "[duration]",
       "limit": 20,
       "offset": 0,
-      "estimatedTotalHits": 4
+      "estimatedTotalHits": 4,
+      "requestUid": "[uuid]"
     }
     "###);
 
@@ -1278,7 +1290,7 @@ async fn federation_sort_same_indexes_same_criterion_same_direction() {
         ]}))
         .await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -1353,7 +1365,8 @@ async fn federation_sort_same_indexes_same_criterion_same_direction() {
       "processingTimeMs": "[duration]",
       "limit": 20,
       "offset": 0,
-      "estimatedTotalHits": 3
+      "estimatedTotalHits": 3,
+      "requestUid": "[uuid]"
     }
     "###);
 }
@@ -1449,7 +1462,7 @@ async fn federation_sort_same_indexes_different_criterion_same_direction() {
         ]}))
         .await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -1538,7 +1551,8 @@ async fn federation_sort_same_indexes_different_criterion_same_direction() {
       "processingTimeMs": "[duration]",
       "limit": 20,
       "offset": 0,
-      "estimatedTotalHits": 4
+      "estimatedTotalHits": 4,
+      "requestUid": "[uuid]"
     }
     "###);
 
@@ -1551,7 +1565,7 @@ async fn federation_sort_same_indexes_different_criterion_same_direction() {
         ]}))
         .await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -1626,7 +1640,8 @@ async fn federation_sort_same_indexes_different_criterion_same_direction() {
       "processingTimeMs": "[duration]",
       "limit": 20,
       "offset": 0,
-      "estimatedTotalHits": 3
+      "estimatedTotalHits": 3,
+      "requestUid": "[uuid]"
     }
     "###);
 }
@@ -1704,7 +1719,7 @@ async fn federation_sort_different_indexes_same_criterion_same_direction() {
         ]}))
         .await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -1831,7 +1846,8 @@ async fn federation_sort_different_indexes_same_criterion_same_direction() {
       "processingTimeMs": "[duration]",
       "limit": 20,
       "offset": 0,
-      "estimatedTotalHits": 10
+      "estimatedTotalHits": 10,
+      "requestUid": "[uuid]"
     }
     "###);
 
@@ -1844,7 +1860,7 @@ async fn federation_sort_different_indexes_same_criterion_same_direction() {
         ]}))
         .await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -1915,7 +1931,8 @@ async fn federation_sort_different_indexes_same_criterion_same_direction() {
       "processingTimeMs": "[duration]",
       "limit": 20,
       "offset": 0,
-      "estimatedTotalHits": 6
+      "estimatedTotalHits": 6,
+      "requestUid": "[uuid]"
     }
     "###);
 }
@@ -1936,7 +1953,7 @@ async fn federation_sort_different_ranking_rules() {
         ]}))
         .await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -2063,7 +2080,8 @@ async fn federation_sort_different_ranking_rules() {
       "processingTimeMs": "[duration]",
       "limit": 20,
       "offset": 0,
-      "estimatedTotalHits": 10
+      "estimatedTotalHits": 10,
+      "requestUid": "[uuid]"
     }
     "###);
 
@@ -2142,7 +2160,7 @@ async fn federation_sort_different_indexes_different_criterion_same_direction()
         ]}))
         .await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -2269,7 +2287,8 @@ async fn federation_sort_different_indexes_different_criterion_same_direction()
       "processingTimeMs": "[duration]",
       "limit": 20,
       "offset": 0,
-      "estimatedTotalHits": 10
+      "estimatedTotalHits": 10,
+      "requestUid": "[uuid]"
     }
     "###);
 
@@ -2282,7 +2301,7 @@ async fn federation_sort_different_indexes_different_criterion_same_direction()
         ]}))
         .await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -2353,7 +2372,8 @@ async fn federation_sort_different_indexes_different_criterion_same_direction()
       "processingTimeMs": "[duration]",
       "limit": 20,
       "offset": 0,
-      "estimatedTotalHits": 6
+      "estimatedTotalHits": 6,
+      "requestUid": "[uuid]"
     }
     "###);
 }
@@ -2424,7 +2444,7 @@ async fn federation_limit_offset() {
             ]}))
             .await;
         snapshot!(code, @"200 OK");
-        snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]" }), @r###"
+        snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]", ".**.requestUid" => "[uuid]" }), @r###"
         {
           "hits": [
             {
@@ -2527,7 +2547,8 @@ async fn federation_limit_offset() {
           "processingTimeMs": "[duration]",
           "limit": 20,
           "offset": 0,
-          "estimatedTotalHits": 12
+          "estimatedTotalHits": 12,
+          "requestUid": "[uuid]"
         }
         "###);
     }
@@ -2549,7 +2570,7 @@ async fn federation_limit_offset() {
             ]}))
             .await;
         snapshot!(code, @"200 OK");
-        snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]" }), @r###"
+        snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]", ".requestUid" => "[uuid]" }), @r###"
         {
           "hits": [
             {
@@ -2564,7 +2585,8 @@ async fn federation_limit_offset() {
           "processingTimeMs": "[duration]",
           "limit": 1,
           "offset": 0,
-          "estimatedTotalHits": 12
+          "estimatedTotalHits": 12,
+          "requestUid": "[uuid]"
         }
         "###);
     }
@@ -2586,7 +2608,7 @@ async fn federation_limit_offset() {
             ]}))
             .await;
         snapshot!(code, @"200 OK");
-        snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]" }), @r###"
+        snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]", ".requestUid" => "[uuid]" }), @r###"
         {
           "hits": [
             {
@@ -2673,7 +2695,8 @@ async fn federation_limit_offset() {
           "processingTimeMs": "[duration]",
           "limit": 20,
           "offset": 2,
-          "estimatedTotalHits": 12
+          "estimatedTotalHits": 12,
+          "requestUid": "[uuid]"
         }
         "###);
     }
@@ -2695,13 +2718,14 @@ async fn federation_limit_offset() {
             ]}))
             .await;
         snapshot!(code, @"200 OK");
-        snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]" }), @r###"
+        snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]", ".requestUid" => "[uuid]" }), @r###"
         {
           "hits": [],
           "processingTimeMs": "[duration]",
           "limit": 20,
           "offset": 12,
-          "estimatedTotalHits": 12
+          "estimatedTotalHits": 12,
+          "requestUid": "[uuid]"
         }
         "###);
     }
@@ -2731,7 +2755,7 @@ async fn federation_formatting() {
             ]}))
             .await;
         snapshot!(code, @"200 OK");
-        snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]" }), @r###"
+        snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]", ".**.requestUid" => "[uuid]" }), @r###"
         {
           "hits": [
             {
@@ -2861,7 +2885,8 @@ async fn federation_formatting() {
           "processingTimeMs": "[duration]",
           "limit": 20,
           "offset": 0,
-          "estimatedTotalHits": 12
+          "estimatedTotalHits": 12,
+          "requestUid": "[uuid]"
         }
         "###);
     }
@@ -2883,7 +2908,7 @@ async fn federation_formatting() {
             ]}))
             .await;
         snapshot!(code, @"200 OK");
-        snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]" }), @r###"
+        snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]", ".requestUid" => "[uuid]" }), @r###"
         {
           "hits": [
             {
@@ -2898,7 +2923,8 @@ async fn federation_formatting() {
           "processingTimeMs": "[duration]",
           "limit": 1,
           "offset": 0,
-          "estimatedTotalHits": 12
+          "estimatedTotalHits": 12,
+          "requestUid": "[uuid]"
         }
         "###);
     }
@@ -2920,7 +2946,7 @@ async fn federation_formatting() {
             ]}))
             .await;
         snapshot!(code, @"200 OK");
-        snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]" }), @r###"
+        snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]", ".requestUid" => "[uuid]" }), @r###"
         {
           "hits": [
             {
@@ -3007,7 +3033,8 @@ async fn federation_formatting() {
           "processingTimeMs": "[duration]",
           "limit": 20,
           "offset": 2,
-          "estimatedTotalHits": 12
+          "estimatedTotalHits": 12,
+          "requestUid": "[uuid]"
         }
         "###);
     }
@@ -3029,13 +3056,14 @@ async fn federation_formatting() {
             ]}))
             .await;
         snapshot!(code, @"200 OK");
-        snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]" }), @r###"
+        snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]", ".requestUid" => "[uuid]" }), @r###"
         {
           "hits": [],
           "processingTimeMs": "[duration]",
           "limit": 20,
           "offset": 12,
-          "estimatedTotalHits": 12
+          "estimatedTotalHits": 12,
+          "requestUid": "[uuid]"
         }
         "###);
     }
@@ -3098,7 +3126,7 @@ async fn federation_null_weight() {
         ]}))
         .await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -3137,7 +3165,8 @@ async fn federation_null_weight() {
       "processingTimeMs": "[duration]",
       "limit": 20,
       "offset": 0,
-      "estimatedTotalHits": 3
+      "estimatedTotalHits": 3,
+      "requestUid": "[uuid]"
     }
     "###);
 }
@@ -3244,7 +3273,7 @@ async fn federation_federated_contains_facets() {
         ]}))
         .await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]", ".**.requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -3280,7 +3309,8 @@ async fn federation_federated_contains_facets() {
       "processingTimeMs": "[duration]",
       "limit": 20,
       "offset": 0,
-      "estimatedTotalHits": 3
+      "estimatedTotalHits": 3,
+      "requestUid": "[uuid]"
     }
     "###);
 
@@ -3488,7 +3518,7 @@ async fn federation_vector_single_index() {
         ]}))
         .await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]", ".**.requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -3532,7 +3562,8 @@ async fn federation_vector_single_index() {
       "limit": 20,
       "offset": 0,
       "estimatedTotalHits": 4,
-      "semanticHitCount": 4
+      "semanticHitCount": 4,
+      "requestUid": "[uuid]"
     }
     "###);
 
@@ -3545,7 +3576,7 @@ async fn federation_vector_single_index() {
         ]}))
         .await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -3589,7 +3620,8 @@ async fn federation_vector_single_index() {
       "limit": 20,
       "offset": 0,
       "estimatedTotalHits": 4,
-      "semanticHitCount": 4
+      "semanticHitCount": 4,
+      "requestUid": "[uuid]"
     }
     "###);
 
@@ -3603,7 +3635,7 @@ async fn federation_vector_single_index() {
         ]}))
         .await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -3651,7 +3683,8 @@ async fn federation_vector_single_index() {
       "limit": 20,
       "offset": 0,
       "estimatedTotalHits": 4,
-      "semanticHitCount": 3
+      "semanticHitCount": 3,
+      "requestUid": "[uuid]"
     }
     "###);
 }
@@ -3703,7 +3736,7 @@ async fn federation_vector_two_indexes() {
         ]}))
         .await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]" }), @r#"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]", ".**.requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -3922,9 +3955,10 @@ async fn federation_vector_two_indexes() {
           0.6
         ]
       },
-      "semanticHitCount": 6
+      "semanticHitCount": 6,
+      "requestUid": "[uuid]"
     }
-    "#);
+    "###);
 
     // hybrid search, distinct embedder
     let (response, code) = server
@@ -3934,7 +3968,7 @@ async fn federation_vector_two_indexes() {
         ]}))
         .await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]" }), @r#"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".**._rankingScore" => "[score]", ".**.requestUid" => "[uuid]" }), @r#"
     {
       "hits": [
         {
@@ -4161,7 +4195,8 @@ async fn federation_vector_two_indexes() {
           0.6
         ]
       },
-      "semanticHitCount": 8
+      "semanticHitCount": 8,
+      "requestUid": "[uuid]"
     }
     "#);
 }
@@ -4209,7 +4244,7 @@ async fn federation_facets_different_indexes_same_facet() {
         ]}))
         .await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -4380,7 +4415,8 @@ async fn federation_facets_different_indexes_same_facet() {
           },
           "stats": {}
         }
-      }
+      },
+      "requestUid": "[uuid]"
     }
     "###);
 
@@ -4399,7 +4435,7 @@ async fn federation_facets_different_indexes_same_facet() {
     ]}))
     .await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -4541,7 +4577,8 @@ async fn federation_facets_different_indexes_same_facet() {
           "Shazam!": 1
         }
       },
-      "facetStats": {}
+      "facetStats": {},
+      "requestUid": "[uuid]"
     }
     "###);
 
@@ -4561,7 +4598,7 @@ async fn federation_facets_different_indexes_same_facet() {
         ]}))
         .await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -4686,7 +4723,8 @@ async fn federation_facets_different_indexes_same_facet() {
           "distribution": {},
           "stats": {}
         }
-      }
+      },
+      "requestUid": "[uuid]"
     }
     "###);
 }
@@ -4748,7 +4786,7 @@ async fn federation_facets_same_indexes() {
         ]}))
         .await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -4806,7 +4844,8 @@ async fn federation_facets_same_indexes() {
             }
           }
         }
-      }
+      },
+      "requestUid": "[uuid]"
     }
     "###);
 
@@ -4822,7 +4861,7 @@ async fn federation_facets_same_indexes() {
         ]}))
         .await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -4908,7 +4947,8 @@ async fn federation_facets_same_indexes() {
             }
           }
         }
-      }
+      },
+      "requestUid": "[uuid]"
     }
     "###);
 
@@ -4925,7 +4965,7 @@ async fn federation_facets_same_indexes() {
         ]}))
         .await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -4987,7 +5027,8 @@ async fn federation_facets_same_indexes() {
           "min": 2.0,
           "max": 6.0
         }
-      }
+      },
+      "requestUid": "[uuid]"
     }
     "###);
 }
@@ -5040,7 +5081,7 @@ async fn federation_inconsistent_merge_order() {
       ]}))
       .await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -5217,7 +5258,8 @@ async fn federation_inconsistent_merge_order() {
           },
           "stats": {}
         }
-      }
+      },
+      "requestUid": "[uuid]"
     }
     "###);
 
@@ -5264,7 +5306,7 @@ async fn federation_inconsistent_merge_order() {
  ]}))
  .await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -5404,7 +5446,8 @@ async fn federation_inconsistent_merge_order() {
           "Batman Returns": 1
         }
       },
-      "facetStats": {}
+      "facetStats": {},
+      "requestUid": "[uuid]"
     }
     "###);
 }

crates/meilisearch/tests/search/multi/proxy.rs

@@ -229,7 +229,7 @@ async fn remote_sharding() {
 
     let (response, _status_code) = ms0.multi_search(request.clone()).await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -287,12 +287,13 @@ async fn remote_sharding() {
       "limit": 20,
       "offset": 0,
       "estimatedTotalHits": 5,
+      "requestUid": "[uuid]",
       "remoteErrors": {}
     }
     "###);
     let (response, _status_code) = ms1.multi_search(request.clone()).await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -350,12 +351,13 @@ async fn remote_sharding() {
       "limit": 20,
       "offset": 0,
       "estimatedTotalHits": 5,
+      "requestUid": "[uuid]",
       "remoteErrors": {}
     }
     "###);
     let (response, _status_code) = ms2.multi_search(request.clone()).await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -413,6 +415,7 @@ async fn remote_sharding() {
       "limit": 20,
       "offset": 0,
       "estimatedTotalHits": 5,
+      "requestUid": "[uuid]",
       "remoteErrors": {}
     }
     "###);
@@ -594,7 +597,7 @@ async fn remote_sharding_retrieve_vectors() {
 
     let (response, _status_code) = ms0.multi_search(request.clone()).await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]" }), @r#"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [],
       "processingTimeMs": "[time]",
@@ -619,9 +622,10 @@ async fn remote_sharding_retrieve_vectors() {
         ]
       },
       "semanticHitCount": 0,
+      "requestUid": "[uuid]",
       "remoteErrors": {}
     }
-    "#);
+    "###);
 
     // multi vector search: two local queries, one remote
 
@@ -669,7 +673,7 @@ async fn remote_sharding_retrieve_vectors() {
 
     let (response, _status_code) = ms0.multi_search(request.clone()).await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]" }), @r#"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]", ".requestUid" => "[uuid]" }), @r#"
     {
       "hits": [],
       "processingTimeMs": "[time]",
@@ -694,6 +698,7 @@ async fn remote_sharding_retrieve_vectors() {
         ]
       },
       "semanticHitCount": 0,
+      "requestUid": "[uuid]",
       "remoteErrors": {}
     }
     "#);
@@ -744,7 +749,7 @@ async fn remote_sharding_retrieve_vectors() {
 
     let (response, _status_code) = ms0.multi_search(request.clone()).await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]" }), @r#"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]", ".requestUid" => "[uuid]" }), @r#"
     {
       "hits": [],
       "processingTimeMs": "[time]",
@@ -769,6 +774,7 @@ async fn remote_sharding_retrieve_vectors() {
         ]
       },
       "semanticHitCount": 0,
+      "requestUid": "[uuid]",
       "remoteErrors": {}
     }
     "#);
@@ -819,7 +825,7 @@ async fn remote_sharding_retrieve_vectors() {
 
     let (response, _status_code) = ms0.multi_search(request.clone()).await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]" }), @r#"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [],
       "processingTimeMs": "[time]",
@@ -839,9 +845,10 @@ async fn remote_sharding_retrieve_vectors() {
         ]
       },
       "semanticHitCount": 0,
+      "requestUid": "[uuid]",
       "remoteErrors": {}
     }
-    "#);
+    "###);
 
     // multi vector search: no local queries, all remote
 
@@ -889,7 +896,7 @@ async fn remote_sharding_retrieve_vectors() {
 
     let (response, _status_code) = ms0.multi_search(request.clone()).await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]" }), @r#"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [],
       "processingTimeMs": "[time]",
@@ -913,9 +920,10 @@ async fn remote_sharding_retrieve_vectors() {
           0.2
         ]
       },
+      "requestUid": "[uuid]",
       "remoteErrors": {}
     }
-    "#);
+    "###);
 }
 
 #[actix_rt::test]
@@ -1133,7 +1141,7 @@ async fn error_no_weighted_score() {
 
     let (response, _status_code) = ms0.multi_search(request.clone()).await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -1161,6 +1169,7 @@ async fn error_no_weighted_score() {
       "limit": 20,
       "offset": 0,
       "estimatedTotalHits": 2,
+      "requestUid": "[uuid]",
       "remoteErrors": {
         "ms1": {
           "message": "remote hit does not contain `._federation.weightedScoreValues`\n  - hint: check that the remote instance is a Meilisearch instance running the same version",
@@ -1272,7 +1281,7 @@ async fn error_bad_response() {
     let (response, _status_code) = ms0.multi_search(request.clone()).await;
     snapshot!(code, @"200 OK");
 
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -1300,6 +1309,7 @@ async fn error_bad_response() {
       "limit": 20,
       "offset": 0,
       "estimatedTotalHits": 2,
+      "requestUid": "[uuid]",
       "remoteErrors": {
         "ms1": {
           "message": "could not parse response from the remote host as a federated search response:\n  - response from remote: <html>Returning an HTML page</html>\n  - hint: check that the remote instance is a Meilisearch instance running the same version",
@@ -1404,7 +1414,7 @@ async fn error_bad_request() {
     let (response, _status_code) = ms0.multi_search(request.clone()).await;
     snapshot!(code, @"200 OK");
 
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -1432,6 +1442,7 @@ async fn error_bad_request() {
       "limit": 20,
       "offset": 0,
       "estimatedTotalHits": 2,
+      "requestUid": "[uuid]",
       "remoteErrors": {
         "ms1": {
           "message": "remote host responded with code 400:\n  - response from remote: {\"message\":\"Inside `.queries[1]`: Index `nottest` not found.\",\"code\":\"index_not_found\",\"type\":\"invalid_request\",\"link\":\"https://docs.meilisearch.com/errors#index_not_found\"}\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",
@@ -1541,7 +1552,7 @@ async fn error_bad_request_facets_by_index() {
     let (response, _status_code) = ms0.multi_search(request.clone()).await;
     snapshot!(code, @"200 OK");
 
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -1575,6 +1586,7 @@ async fn error_bad_request_facets_by_index() {
           "stats": {}
         }
       },
+      "requestUid": "[uuid]",
       "remoteErrors": {
         "ms1": {
           "message": "remote host responded with code 400:\n  - response from remote: {\"message\":\"Inside `.federation.facetsByIndex.test0`: Index `test0` not found.\\n - Note: index `test0` is not used in queries\",\"code\":\"index_not_found\",\"type\":\"invalid_request\",\"link\":\"https://docs.meilisearch.com/errors#index_not_found\"}\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",
@@ -1687,7 +1699,7 @@ async fn error_bad_request_facets_by_index_facet() {
     let (response, _status_code) = ms0.multi_search(request.clone()).await;
     snapshot!(code, @"200 OK");
 
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -1726,6 +1738,7 @@ async fn error_bad_request_facets_by_index_facet() {
           "stats": {}
         }
       },
+      "requestUid": "[uuid]",
       "remoteErrors": {
         "ms1": {
           "message": "remote host responded with code 400:\n  - response from remote: {\"message\":\"Inside `.federation.facetsByIndex.test`: Invalid facet distribution: Attribute `id` is not filterable. This index does not have configured filterable attributes.\\n - Note: index `test` used in `.queries[1]`\",\"code\":\"invalid_multi_search_facets\",\"type\":\"invalid_request\",\"link\":\"https://docs.meilisearch.com/errors#invalid_multi_search_facets\"}\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",
@@ -2035,7 +2048,7 @@ async fn error_remote_404() {
 
     let (response, _status_code) = ms0.multi_search(request.clone()).await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -2063,6 +2076,7 @@ async fn error_remote_404() {
       "limit": 20,
       "offset": 0,
       "estimatedTotalHits": 2,
+      "requestUid": "[uuid]",
       "remoteErrors": {
         "ms1": {
           "message": "remote host responded with code 404:\n  - response from remote: null\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",
@@ -2075,7 +2089,7 @@ async fn error_remote_404() {
     "###);
     let (response, _status_code) = ms1.multi_search(request.clone()).await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -2113,6 +2127,7 @@ async fn error_remote_404() {
       "limit": 20,
       "offset": 0,
       "estimatedTotalHits": 3,
+      "requestUid": "[uuid]",
       "remoteErrors": {}
     }
     "###);
@@ -2244,7 +2259,7 @@ async fn error_remote_sharding_auth() {
 
     let (response, _status_code) = ms0.multi_search(request.clone()).await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -2272,6 +2287,7 @@ async fn error_remote_sharding_auth() {
       "limit": 20,
       "offset": 0,
       "estimatedTotalHits": 2,
+      "requestUid": "[uuid]",
       "remoteErrors": {
         "ms1-notsearch": {
           "message": "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",
@@ -2405,7 +2421,7 @@ async fn remote_sharding_auth() {
 
     let (response, _status_code) = ms0.multi_search(request.clone()).await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -2453,6 +2469,7 @@ async fn remote_sharding_auth() {
       "limit": 20,
       "offset": 0,
       "estimatedTotalHits": 4,
+      "requestUid": "[uuid]",
       "remoteErrors": {}
     }
     "###);
@@ -2555,7 +2572,7 @@ async fn error_remote_500() {
 
     let (response, _status_code) = ms0.multi_search(request.clone()).await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -2583,6 +2600,7 @@ async fn error_remote_500() {
       "limit": 20,
       "offset": 0,
       "estimatedTotalHits": 2,
+      "requestUid": "[uuid]",
       "remoteErrors": {
         "ms1": {
           "message": "remote host responded with code 500:\n  - response from remote: {\"error\":\"provoked error\",\"code\":\"test_error\",\"link\":\"https://docs.meilisearch.com/errors#test_error\"}",
@@ -2596,7 +2614,7 @@ async fn error_remote_500() {
     let (response, _status_code) = ms1.multi_search(request.clone()).await;
     snapshot!(code, @"200 OK");
     // the response if full because we queried the instance that works
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -2634,6 +2652,7 @@ async fn error_remote_500() {
       "limit": 20,
       "offset": 0,
       "estimatedTotalHits": 3,
+      "requestUid": "[uuid]",
       "remoteErrors": {}
     }
     "###);
@@ -2737,7 +2756,7 @@ async fn error_remote_500_once() {
     // Meilisearch is tolerant to a single failure
     let (response, _status_code) = ms0.multi_search(request.clone()).await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -2775,12 +2794,13 @@ async fn error_remote_500_once() {
       "limit": 20,
       "offset": 0,
       "estimatedTotalHits": 3,
+      "requestUid": "[uuid]",
       "remoteErrors": {}
     }
     "###);
     let (response, _status_code) = ms1.multi_search(request.clone()).await;
     snapshot!(code, @"200 OK");
-    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]" }), @r###"
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -2818,6 +2838,7 @@ async fn error_remote_500_once() {
       "limit": 20,
       "offset": 0,
       "estimatedTotalHits": 3,
+      "requestUid": "[uuid]",
       "remoteErrors": {}
     }
     "###);
@@ -2845,7 +2866,8 @@ async fn error_remote_timeout() {
     snapshot!(json_string!(response), @r###"
     {
       "self": "ms0",
-      "remotes": {}
+      "remotes": {},
+      "sharding": false
     }
     "###);
     let (response, code) = ms1.set_network(json!({"self": "ms1"})).await;
@@ -2853,7 +2875,8 @@ async fn error_remote_timeout() {
     snapshot!(json_string!(response), @r###"
     {
       "self": "ms1",
-      "remotes": {}
+      "remotes": {},
+      "sharding": false
     }
     "###);
 
@@ -2874,7 +2897,7 @@ async fn error_remote_timeout() {
     let rms0 = LocalMeili::new(ms0.clone()).await;
     let rms1 = LocalMeili::with_params(
         ms1.clone(),
-        LocalMeiliParams { delay: Some(std::time::Duration::from_secs(6)), ..Default::default() },
+        LocalMeiliParams { delay: Some(std::time::Duration::from_secs(31)), ..Default::default() },
     )
     .await;
 

crates/meilisearch/tests/search/pagination.rs

@@ -175,7 +175,8 @@ async fn test_issue_5274() {
 
     snapshot!(json_string!(rep, {
         ".processingTimeMs" => "[ignored]",
-    }), @r#"
+        ".requestUid" => "[uuid]"
+    }), @r###"
     {
       "hits": [
         {
@@ -189,7 +190,8 @@ async fn test_issue_5274() {
       "hitsPerPage": 1,
       "page": 1,
       "totalPages": 1,
-      "totalHits": 1
+      "totalHits": 1,
+      "requestUid": "[uuid]"
     }
-    "#);
+    "###);
 }

crates/meilisearch/tests/tasks/errors.rs

@@ -97,7 +97,7 @@ async fn task_bad_types() {
     snapshot!(code, @"400 Bad Request");
     snapshot!(json_string!(response), @r#"
     {
-      "message": "Invalid value in parameter `types`: `doggo` is not a valid task type. Available types are `documentAdditionOrUpdate`, `documentEdition`, `documentDeletion`, `settingsUpdate`, `indexCreation`, `indexDeletion`, `indexUpdate`, `indexSwap`, `taskCancelation`, `taskDeletion`, `dumpCreation`, `snapshotCreation`, `export`, `upgradeDatabase`.",
+      "message": "Invalid value in parameter `types`: `doggo` is not a valid task type. Available types are `documentAdditionOrUpdate`, `documentEdition`, `documentDeletion`, `settingsUpdate`, `indexCreation`, `indexDeletion`, `indexUpdate`, `indexSwap`, `taskCancelation`, `taskDeletion`, `dumpCreation`, `snapshotCreation`, `export`, `upgradeDatabase`, `indexCompaction`.",
       "code": "invalid_task_types",
       "type": "invalid_request",
       "link": "https://docs.meilisearch.com/errors#invalid_task_types"
@@ -108,7 +108,7 @@ async fn task_bad_types() {
     snapshot!(code, @"400 Bad Request");
     snapshot!(json_string!(response), @r#"
     {
-      "message": "Invalid value in parameter `types`: `doggo` is not a valid task type. Available types are `documentAdditionOrUpdate`, `documentEdition`, `documentDeletion`, `settingsUpdate`, `indexCreation`, `indexDeletion`, `indexUpdate`, `indexSwap`, `taskCancelation`, `taskDeletion`, `dumpCreation`, `snapshotCreation`, `export`, `upgradeDatabase`.",
+      "message": "Invalid value in parameter `types`: `doggo` is not a valid task type. Available types are `documentAdditionOrUpdate`, `documentEdition`, `documentDeletion`, `settingsUpdate`, `indexCreation`, `indexDeletion`, `indexUpdate`, `indexSwap`, `taskCancelation`, `taskDeletion`, `dumpCreation`, `snapshotCreation`, `export`, `upgradeDatabase`, `indexCompaction`.",
       "code": "invalid_task_types",
       "type": "invalid_request",
       "link": "https://docs.meilisearch.com/errors#invalid_task_types"
@@ -119,7 +119,7 @@ async fn task_bad_types() {
     snapshot!(code, @"400 Bad Request");
     snapshot!(json_string!(response), @r#"
     {
-      "message": "Invalid value in parameter `types`: `doggo` is not a valid task type. Available types are `documentAdditionOrUpdate`, `documentEdition`, `documentDeletion`, `settingsUpdate`, `indexCreation`, `indexDeletion`, `indexUpdate`, `indexSwap`, `taskCancelation`, `taskDeletion`, `dumpCreation`, `snapshotCreation`, `export`, `upgradeDatabase`.",
+      "message": "Invalid value in parameter `types`: `doggo` is not a valid task type. Available types are `documentAdditionOrUpdate`, `documentEdition`, `documentDeletion`, `settingsUpdate`, `indexCreation`, `indexDeletion`, `indexUpdate`, `indexSwap`, `taskCancelation`, `taskDeletion`, `dumpCreation`, `snapshotCreation`, `export`, `upgradeDatabase`, `indexCompaction`.",
       "code": "invalid_task_types",
       "type": "invalid_request",
       "link": "https://docs.meilisearch.com/errors#invalid_task_types"

crates/meilisearch/tests/upgrade/mod.rs

@@ -43,7 +43,7 @@ async fn version_too_old() {
     std::fs::write(db_path.join("VERSION"), "1.11.9999").unwrap();
     let options = Opt { experimental_dumpless_upgrade: true, ..default_settings };
     let err = Server::new_with_options(options).await.map(|_| ()).unwrap_err();
-    snapshot!(err, @"Database version 1.11.9999 is too old for the experimental dumpless upgrade feature. Please generate a dump using the v1.11.9999 and import it in the v1.22.1");
+    snapshot!(err, @"Database version 1.11.9999 is too old for the experimental dumpless upgrade feature. Please generate a dump using the v1.11.9999 and import it in the v1.24.0");
 }
 
 #[actix_rt::test]
@@ -58,7 +58,7 @@ async fn version_requires_downgrade() {
     std::fs::write(db_path.join("VERSION"), format!("{major}.{minor}.{patch}")).unwrap();
     let options = Opt { experimental_dumpless_upgrade: true, ..default_settings };
     let err = Server::new_with_options(options).await.map(|_| ()).unwrap_err();
-    snapshot!(err, @"Database version 1.22.2 is higher than the Meilisearch version 1.22.1. Downgrade is not supported");
+    snapshot!(err, @"Database version 1.24.1 is higher than the Meilisearch version 1.24.0. Downgrade is not supported");
 }
 
 #[actix_rt::test]

crates/meilisearch/tests/upgrade/v1_12/snapshots/v1_12_0.rs/check_the_index_features/search_with_sort_and_filter.snap

@@ -1,6 +1,5 @@
 ---
 source: crates/meilisearch/tests/upgrade/v1_12/v1_12_0.rs
-snapshot_kind: text
 ---
 {
   "hits": [
@@ -21,5 +20,6 @@ snapshot_kind: text
   "processingTimeMs": "[duration]",
   "limit": 20,
   "offset": 0,
-  "estimatedTotalHits": 1
+  "estimatedTotalHits": 1,
+  "requestUid": "[uuid]"
 }

crates/meilisearch/tests/upgrade/v1_12/snapshots/v1_12_0.rs/check_the_index_scheduler/batches_filter_afterEnqueuedAt_equal_2025-01-16T16_47_41.snap

@@ -8,7 +8,7 @@ source: crates/meilisearch/tests/upgrade/v1_12/v1_12_0.rs
       "progress": null,
       "details": {
         "upgradeFrom": "v1.12.0",
-        "upgradeTo": "v1.22.1"
+        "upgradeTo": "v1.24.0"
       },
       "stats": {
         "totalNbTasks": 1,

crates/meilisearch/tests/upgrade/v1_12/snapshots/v1_12_0.rs/check_the_index_scheduler/batches_filter_afterFinishedAt_equal_2025-01-16T16_47_41.snap

@@ -8,7 +8,7 @@ source: crates/meilisearch/tests/upgrade/v1_12/v1_12_0.rs
       "progress": null,
       "details": {
         "upgradeFrom": "v1.12.0",
-        "upgradeTo": "v1.22.1"
+        "upgradeTo": "v1.24.0"
       },
       "stats": {
         "totalNbTasks": 1,

crates/meilisearch/tests/upgrade/v1_12/snapshots/v1_12_0.rs/check_the_index_scheduler/batches_filter_afterStartedAt_equal_2025-01-16T16_47_41.snap

@@ -8,7 +8,7 @@ source: crates/meilisearch/tests/upgrade/v1_12/v1_12_0.rs
       "progress": null,
       "details": {
         "upgradeFrom": "v1.12.0",
-        "upgradeTo": "v1.22.1"
+        "upgradeTo": "v1.24.0"
       },
       "stats": {
         "totalNbTasks": 1,

crates/meilisearch/tests/upgrade/v1_12/snapshots/v1_12_0.rs/check_the_index_scheduler/tasks_filter_afterEnqueuedAt_equal_2025-01-16T16_47_41.snap

@@ -12,7 +12,7 @@ source: crates/meilisearch/tests/upgrade/v1_12/v1_12_0.rs
       "canceledBy": null,
       "details": {
         "upgradeFrom": "v1.12.0",
-        "upgradeTo": "v1.22.1"
+        "upgradeTo": "v1.24.0"
       },
       "error": null,
       "duration": "[duration]",

crates/meilisearch/tests/upgrade/v1_12/snapshots/v1_12_0.rs/check_the_index_scheduler/tasks_filter_afterFinishedAt_equal_2025-01-16T16_47_41.snap

@@ -12,7 +12,7 @@ source: crates/meilisearch/tests/upgrade/v1_12/v1_12_0.rs
       "canceledBy": null,
       "details": {
         "upgradeFrom": "v1.12.0",
-        "upgradeTo": "v1.22.1"
+        "upgradeTo": "v1.24.0"
       },
       "error": null,
       "duration": "[duration]",

crates/meilisearch/tests/upgrade/v1_12/snapshots/v1_12_0.rs/check_the_index_scheduler/tasks_filter_afterStartedAt_equal_2025-01-16T16_47_41.snap

@@ -12,7 +12,7 @@ source: crates/meilisearch/tests/upgrade/v1_12/v1_12_0.rs
       "canceledBy": null,
       "details": {
         "upgradeFrom": "v1.12.0",
-        "upgradeTo": "v1.22.1"
+        "upgradeTo": "v1.24.0"
       },
       "error": null,
       "duration": "[duration]",

crates/meilisearch/tests/upgrade/v1_12/snapshots/v1_12_0.rs/check_the_index_scheduler/the_whole_batch_queue_once_everything_has_been_processed.snap

@@ -8,7 +8,7 @@ source: crates/meilisearch/tests/upgrade/v1_12/v1_12_0.rs
       "progress": null,
       "details": {
         "upgradeFrom": "v1.12.0",
-        "upgradeTo": "v1.22.1"
+        "upgradeTo": "v1.24.0"
       },
       "stats": {
         "totalNbTasks": 1,

crates/meilisearch/tests/upgrade/v1_12/snapshots/v1_12_0.rs/check_the_index_scheduler/the_whole_task_queue_once_everything_has_been_processed.snap

@@ -12,7 +12,7 @@ source: crates/meilisearch/tests/upgrade/v1_12/v1_12_0.rs
       "canceledBy": null,
       "details": {
         "upgradeFrom": "v1.12.0",
-        "upgradeTo": "v1.22.1"
+        "upgradeTo": "v1.24.0"
       },
       "error": null,
       "duration": "[duration]",

crates/meilisearch/tests/upgrade/v1_12/v1_12_0.rs

@@ -294,7 +294,7 @@ async fn check_the_index_features(server: &Server) {
 
     let (results, _status) =
         kefir.search_post(json!({ "sort": ["age:asc"], "filter": "surname = kefirounet" })).await;
-    snapshot!(results, name: "search_with_sort_and_filter");
+    snapshot!(json_string!(results, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), name: "search_with_sort_and_filter");
 
     // ensuring we can get the vectors and their `regenerate` is still good.
     let (results, _status) = kefir.search_post(json!({"retrieveVectors": true})).await;

crates/meilisearch/tests/vector/binary_quantized.rs

@@ -323,7 +323,7 @@ async fn binary_quantize_clear_documents() {
     // Make sure the vector DB has been cleared
     let (documents, _code) =
         index.search_post(json!({ "hybrid": { "embedder": "manual" }, "vector": [1, 1, 1] })).await;
-    snapshot!(documents, @r#"
+    snapshot!(json_string!(documents, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [],
       "query": "",
@@ -331,9 +331,10 @@ async fn binary_quantize_clear_documents() {
       "limit": 20,
       "offset": 0,
       "estimatedTotalHits": 0,
+      "requestUid": "[uuid]",
       "semanticHitCount": 0
     }
-    "#);
+    "###);
 }
 
 #[actix_rt::test]

crates/meilisearch/tests/vector/fragments.rs

@@ -257,7 +257,7 @@ async fn search_with_vector() {
         json!({"vector": [1.0, 1.0, 1.0], "hybrid": {"semanticRatio": 1.0, "embedder": "rest"}, "limit": 1}
     )).await;
     snapshot!(code, @"200 OK");
-    snapshot!(value, @r#"
+    snapshot!(json_string!(value, { ".requestUid" => "[uuid]", ".processingTimeMs" => "[duration]" }), @r###"
     {
       "hits": [
         {
@@ -270,9 +270,10 @@ async fn search_with_vector() {
       "limit": 1,
       "offset": 0,
       "estimatedTotalHits": 4,
+      "requestUid": "[uuid]",
       "semanticHitCount": 1
     }
-    "#);
+    "###);
 }
 
 #[actix_rt::test]
@@ -288,7 +289,7 @@ async fn search_with_media() {
         ))
         .await;
     snapshot!(code, @"200 OK");
-    snapshot!(value, @r#"
+    snapshot!(json_string!(value, { ".requestUid" => "[uuid]", ".processingTimeMs" => "[duration]" }), @r###"
     {
       "hits": [
         {
@@ -302,9 +303,10 @@ async fn search_with_media() {
       "limit": 1,
       "offset": 0,
       "estimatedTotalHits": 4,
+      "requestUid": "[uuid]",
       "semanticHitCount": 1
     }
-    "#);
+    "###);
 }
 
 #[actix_rt::test]
@@ -390,7 +392,7 @@ async fn search_with_query() {
         ))
         .await;
     snapshot!(code, @"200 OK");
-    snapshot!(value, @r#"
+    snapshot!(json_string!(value, { ".requestUid" => "[uuid]", ".processingTimeMs" => "[duration]" }), @r###"
     {
       "hits": [
         {
@@ -404,9 +406,10 @@ async fn search_with_query() {
       "limit": 1,
       "offset": 0,
       "estimatedTotalHits": 4,
+      "requestUid": "[uuid]",
       "semanticHitCount": 1
     }
-    "#);
+    "###);
 }
 
 #[actix_rt::test]
@@ -2076,7 +2079,7 @@ async fn composite() {
         json!({"vector": [1.0, 1.0, 1.0], "hybrid": {"semanticRatio": 1.0, "embedder": "rest"}, "limit": 1}
     )).await;
     snapshot!(code, @"200 OK");
-    snapshot!(value, @r#"
+    snapshot!(json_string!(value, { ".requestUid" => "[uuid]", ".processingTimeMs" => "[duration]" }), @r###"
     {
       "hits": [
         {
@@ -2089,9 +2092,10 @@ async fn composite() {
       "limit": 1,
       "offset": 0,
       "estimatedTotalHits": 4,
+      "requestUid": "[uuid]",
       "semanticHitCount": 1
     }
-    "#);
+    "###);
 
     let (value, code) = index
         .search_post(
@@ -2100,7 +2104,7 @@ async fn composite() {
         )
         .await;
     snapshot!(code, @"200 OK");
-    snapshot!(value, @r#"
+    snapshot!(json_string!(value, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r#"
     {
       "hits": [
         {
@@ -2114,6 +2118,7 @@ async fn composite() {
       "limit": 1,
       "offset": 0,
       "estimatedTotalHits": 4,
+      "requestUid": "[uuid]",
       "semanticHitCount": 1
     }
     "#);

crates/meilisearch/tests/vector/mod.rs

@@ -689,7 +689,7 @@ async fn clear_documents() {
     // Make sure the vector DB has been cleared
     let (documents, _code) =
         index.search_post(json!({ "vector": [1, 1, 1], "hybrid": {"embedder": "manual"} })).await;
-    snapshot!(documents, @r#"
+    snapshot!(json_string!(documents, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [],
       "query": "",
@@ -697,9 +697,10 @@ async fn clear_documents() {
       "limit": 20,
       "offset": 0,
       "estimatedTotalHits": 0,
+      "requestUid": "[uuid]",
       "semanticHitCount": 0
     }
-    "#);
+    "###);
 }
 
 #[actix_rt::test]
@@ -743,7 +744,7 @@ async fn add_remove_one_vector_4588() {
             json!({"vector": [1, 1, 1], "hybrid": {"semanticRatio": 1.0, "embedder": "manual"} }),
         )
         .await;
-    snapshot!(documents, @r#"
+    snapshot!(json_string!(documents, { ".processingTimeMs" => "[duration]", ".requestUid" => "[uuid]" }), @r###"
     {
       "hits": [
         {
@@ -756,9 +757,10 @@ async fn add_remove_one_vector_4588() {
       "limit": 20,
       "offset": 0,
       "estimatedTotalHits": 1,
+      "requestUid": "[uuid]",
       "semanticHitCount": 1
     }
-    "#);
+    "###);
 
     let (documents, _code) = index
         .get_all_documents(GetAllDocumentsOptions { retrieve_vectors: true, ..Default::default() })

crates/meilitool/src/main.rs

@@ -126,7 +126,7 @@ enum Command {
     /// before running the copy and compaction. This way the current indexation must finish before
     /// the compaction operation can start. Once the compaction is done, the big index is replaced
     /// by the compacted one and the mutable transaction is released.
-    CompactIndex { index_name: String },
+    IndexCompaction { index_name: String },
 
     /// Uses the hair dryer the dedicate pages hot in cache
     ///
@@ -165,7 +165,7 @@ fn main() -> anyhow::Result<()> {
             let target_version = parse_version(&target_version).context("While parsing `--target-version`. Make sure `--target-version` is in the format MAJOR.MINOR.PATCH")?;
             OfflineUpgrade { db_path, current_version: detected_version, target_version }.upgrade()
         }
-        Command::CompactIndex { index_name } => compact_index(db_path, &index_name),
+        Command::IndexCompaction { index_name } => compact_index(db_path, &index_name),
         Command::HairDryer { index_name, index_part } => {
             hair_dryer(db_path, &index_name, &index_part)
         }

crates/milli/Cargo.toml

@@ -19,7 +19,7 @@ bstr = "1.12.0"
 bytemuck = { version = "1.23.1", features = ["extern_crate_alloc"] }
 byteorder = "1.5.0"
 charabia = { version = "0.9.7", default-features = false }
-cellulite = "0.3.0"
+cellulite = "0.3.1-nested-rtxns-2"
 concat-arrays = "0.1.2"
 convert_case = "0.8.0"
 crossbeam-channel = "0.5.15"
@@ -34,7 +34,7 @@ grenad = { version = "0.5.0", default-features = false, features = [
     "rayon",
     "tempfile",
 ] }
-heed = { version = "0.22.0", default-features = false, features = [
+heed = { version = "0.22.1-nested-rtxns", default-features = false, features = [
     "serde-json",
     "serde-bincode",
 ] }
@@ -89,8 +89,8 @@ rhai = { version = "1.22.2", features = [
     "no_time",
     "sync",
 ] }
-arroy = "0.6.3"
-hannoy = { version = "0.0.8", features = ["arroy"] }
+arroy = "0.6.4-nested-rtxns"
+hannoy = { version = "0.0.9-nested-rtxns-2", features = ["arroy"] }
 rand = "0.8.5"
 tracing = "0.1.41"
 ureq = { version = "2.12.1", features = ["json"] }

crates/milli/src/index.rs

@@ -1983,11 +1983,6 @@ impl Index {
 
         Ok(sizes)
     }
-
-    /// The underlying env for raw access
-    pub fn raw_env(&self) -> &heed::Env<WithoutTls> {
-        &self.env
-    }
 }
 
 pub struct EmbeddingsWithMetadata {

crates/milli/src/score_details.rs

@@ -66,15 +66,29 @@ impl ScoreDetails {
         }
     }
 
+    /// Calculate the global score of the details.
+    ///
+    /// It is computed from the ranks of the ranking rules, excluding the sort/geo sort rules.
+    /// If the details contain a semantic score (ScoreDetails::Vector), it is used instead of the ranking score.
+    ///
+    /// note: this function expects a maximum of one semantic score, otherwise only the last one will be used.
     pub fn global_score<'a>(details: impl Iterator<Item = &'a Self> + 'a) -> f64 {
-        Self::score_values(details)
-            .find_map(|x| {
-                let ScoreValue::Score(score) = x else {
-                    return None;
-                };
-                Some(score)
-            })
-            .unwrap_or(1.0f64)
+        // Filter out only the ranking scores (Rank values) and exclude sort/geo sort
+        let mut semantic_score = None;
+        let ranking_ranks = details.filter_map(|detail| match detail.rank_or_value() {
+            RankOrValue::Rank(rank) => Some(rank),
+            RankOrValue::Score(score) => {
+                semantic_score = Some(score);
+                None
+            }
+            RankOrValue::Sort(_) => None,
+            RankOrValue::GeoSort(_) => None,
+        });
+
+        let ranking_score = Rank::global_score(ranking_ranks);
+
+        // If we have semantic score, use it, otherwise use ranking score
+        semantic_score.unwrap_or(ranking_score)
     }
 
     pub fn score_values<'a>(

crates/milli/src/search/new/bucket_sort.rs

@@ -97,7 +97,7 @@ pub fn bucket_sort<'ctx, Q: RankingRuleQueryTrait>(
 
     logger.start_iteration_ranking_rule(0, ranking_rules[0].as_ref(), query, universe);
 
-    ranking_rules[0].start_iteration(ctx, logger, universe, query)?;
+    ranking_rules[0].start_iteration(ctx, logger, universe, query, &time_budget)?;
 
     let mut ranking_rule_scores: Vec<ScoreDetails> = vec![];
 
@@ -168,42 +168,6 @@ pub fn bucket_sort<'ctx, Q: RankingRuleQueryTrait>(
         };
 
     while valid_docids.len() < max_len_to_evaluate {
-        if time_budget.exceeded() {
-            loop {
-                let bucket = std::mem::take(&mut ranking_rule_universes[cur_ranking_rule_index]);
-                ranking_rule_scores.push(ScoreDetails::Skipped);
-
-                // remove candidates from the universe without adding them to result if their score is below the threshold
-                let is_below_threshold =
-                    ranking_score_threshold.is_some_and(|ranking_score_threshold| {
-                        let current_score = ScoreDetails::global_score(ranking_rule_scores.iter());
-                        current_score < ranking_score_threshold
-                    });
-
-                if is_below_threshold {
-                    all_candidates -= &bucket;
-                    all_candidates -= &ranking_rule_universes[cur_ranking_rule_index];
-                } else {
-                    maybe_add_to_results!(bucket);
-                }
-
-                ranking_rule_scores.pop();
-
-                if cur_ranking_rule_index == 0 {
-                    break;
-                }
-
-                back!();
-            }
-
-            return Ok(BucketSortOutput {
-                scores: valid_scores,
-                docids: valid_docids,
-                all_candidates,
-                degraded: true,
-            });
-        }
-
         // The universe for this bucket is zero, so we don't need to sort
         // anything, just go back to the parent ranking rule.
         if ranking_rule_universes[cur_ranking_rule_index].is_empty()
@@ -216,14 +180,63 @@ pub fn bucket_sort<'ctx, Q: RankingRuleQueryTrait>(
             continue;
         }
 
-        let Some(next_bucket) = ranking_rules[cur_ranking_rule_index].next_bucket(
-            ctx,
-            logger,
-            &ranking_rule_universes[cur_ranking_rule_index],
-        )?
-        else {
-            back!();
-            continue;
+        let next_bucket = if time_budget.exceeded() {
+            match ranking_rules[cur_ranking_rule_index].non_blocking_next_bucket(
+                ctx,
+                logger,
+                &ranking_rule_universes[cur_ranking_rule_index],
+            )? {
+                std::task::Poll::Ready(bucket) => bucket,
+                std::task::Poll::Pending => {
+                    loop {
+                        let bucket =
+                            std::mem::take(&mut ranking_rule_universes[cur_ranking_rule_index]);
+                        ranking_rule_scores.push(ScoreDetails::Skipped);
+
+                        // remove candidates from the universe without adding them to result if their score is below the threshold
+                        let is_below_threshold =
+                            ranking_score_threshold.is_some_and(|ranking_score_threshold| {
+                                let current_score =
+                                    ScoreDetails::global_score(ranking_rule_scores.iter());
+                                current_score < ranking_score_threshold
+                            });
+
+                        if is_below_threshold {
+                            all_candidates -= &bucket;
+                            all_candidates -= &ranking_rule_universes[cur_ranking_rule_index];
+                        } else {
+                            maybe_add_to_results!(bucket);
+                        }
+
+                        ranking_rule_scores.pop();
+
+                        if cur_ranking_rule_index == 0 {
+                            break;
+                        }
+
+                        back!();
+                    }
+
+                    return Ok(BucketSortOutput {
+                        scores: valid_scores,
+                        docids: valid_docids,
+                        all_candidates,
+                        degraded: true,
+                    });
+                }
+            }
+        } else {
+            let Some(next_bucket) = ranking_rules[cur_ranking_rule_index].next_bucket(
+                ctx,
+                logger,
+                &ranking_rule_universes[cur_ranking_rule_index],
+                &time_budget,
+            )?
+            else {
+                back!();
+                continue;
+            };
+            next_bucket
         };
 
         ranking_rule_scores.push(next_bucket.score);
@@ -275,6 +288,7 @@ pub fn bucket_sort<'ctx, Q: RankingRuleQueryTrait>(
             logger,
             &next_bucket.candidates,
             &next_bucket.query,
+            &time_budget,
         )?;
     }
 

crates/milli/src/search/new/exact_attribute.rs

@@ -6,7 +6,7 @@ use super::ranking_rules::{RankingRule, RankingRuleOutput};
 use crate::score_details::{self, ScoreDetails};
 use crate::search::new::query_graph::QueryNodeData;
 use crate::search::new::query_term::ExactTerm;
-use crate::{CboRoaringBitmapCodec, Result, SearchContext, SearchLogger};
+use crate::{CboRoaringBitmapCodec, Result, SearchContext, SearchLogger, TimeBudget};
 
 /// A ranking rule that produces 3 disjoint buckets:
 ///
@@ -35,6 +35,7 @@ impl<'ctx> RankingRule<'ctx, QueryGraph> for ExactAttribute {
         _logger: &mut dyn SearchLogger<QueryGraph>,
         universe: &roaring::RoaringBitmap,
         query: &QueryGraph,
+        _time_budget: &TimeBudget,
     ) -> Result<()> {
         self.state = State::start_iteration(ctx, universe, query)?;
         Ok(())
@@ -46,6 +47,7 @@ impl<'ctx> RankingRule<'ctx, QueryGraph> for ExactAttribute {
         _ctx: &mut SearchContext<'ctx>,
         _logger: &mut dyn SearchLogger<QueryGraph>,
         universe: &roaring::RoaringBitmap,
+        _time_budget: &TimeBudget,
     ) -> Result<Option<RankingRuleOutput<QueryGraph>>> {
         let state = std::mem::take(&mut self.state);
         let (state, output) = State::next(state, universe);

crates/milli/src/search/new/geo_sort.rs

@@ -7,7 +7,7 @@ use super::ranking_rules::{RankingRule, RankingRuleOutput, RankingRuleQueryTrait
 use crate::documents::geo_sort::{fill_cache, next_bucket};
 use crate::documents::{GeoSortParameter, GeoSortStrategy};
 use crate::score_details::{self, ScoreDetails};
-use crate::{GeoPoint, Result, SearchContext, SearchLogger};
+use crate::{GeoPoint, Result, SearchContext, SearchLogger, TimeBudget};
 
 pub struct GeoSort<Q: RankingRuleQueryTrait> {
     query: Option<Q>,
@@ -84,6 +84,7 @@ impl<'ctx, Q: RankingRuleQueryTrait> RankingRule<'ctx, Q> for GeoSort<Q> {
         _logger: &mut dyn SearchLogger<Q>,
         universe: &RoaringBitmap,
         query: &Q,
+        _time_budget: &TimeBudget,
     ) -> Result<()> {
         assert!(self.query.is_none());
 
@@ -110,6 +111,7 @@ impl<'ctx, Q: RankingRuleQueryTrait> RankingRule<'ctx, Q> for GeoSort<Q> {
         ctx: &mut SearchContext<'ctx>,
         _logger: &mut dyn SearchLogger<Q>,
         universe: &RoaringBitmap,
+        _time_budget: &TimeBudget,
     ) -> Result<Option<RankingRuleOutput<Q>>> {
         let query = self.query.as_ref().unwrap().clone();
 

crates/milli/src/search/new/graph_based_ranking_rule.rs

@@ -53,7 +53,7 @@ use super::{QueryGraph, RankingRule, RankingRuleOutput, SearchContext};
 use crate::score_details::Rank;
 use crate::search::new::query_term::LocatedQueryTermSubset;
 use crate::search::new::ranking_rule_graph::PathVisitor;
-use crate::{Result, TermsMatchingStrategy};
+use crate::{Result, TermsMatchingStrategy, TimeBudget};
 
 pub type Words = GraphBasedRankingRule<WordsGraph>;
 impl GraphBasedRankingRule<WordsGraph> {
@@ -135,6 +135,7 @@ impl<'ctx, G: RankingRuleGraphTrait> RankingRule<'ctx, QueryGraph> for GraphBase
         _logger: &mut dyn SearchLogger<QueryGraph>,
         _universe: &RoaringBitmap,
         query_graph: &QueryGraph,
+        _time_budget: &TimeBudget,
     ) -> Result<()> {
         // the `next_max_cost` is the successor integer to the maximum cost of the paths in the graph.
         //
@@ -217,6 +218,7 @@ impl<'ctx, G: RankingRuleGraphTrait> RankingRule<'ctx, QueryGraph> for GraphBase
         ctx: &mut SearchContext<'ctx>,
         logger: &mut dyn SearchLogger<QueryGraph>,
         universe: &RoaringBitmap,
+        _time_budget: &TimeBudget,
     ) -> Result<Option<RankingRuleOutput<QueryGraph>>> {
         // Will crash if `next_bucket` is called before `start_iteration` or after `end_iteration`,
         // should never happen

crates/milli/src/search/new/mod.rs

@@ -21,7 +21,7 @@ mod vector_sort;
 #[cfg(test)]
 mod tests;
 
-use std::collections::HashSet;
+use std::collections::{HashMap, HashSet};
 use std::ops::AddAssign;
 use std::time::Duration;
 
@@ -64,6 +64,12 @@ use crate::{
     UserError, Weight,
 };
 
+/// Cache for synonyms to avoid repeated database access
+#[derive(Default)]
+pub struct SynonymCache {
+    pub cache: Option<HashMap<Vec<String>, Vec<Vec<String>>>>,
+}
+
 /// A structure used throughout the execution of a search query.
 pub struct SearchContext<'ctx> {
     pub index: &'ctx Index,
@@ -73,6 +79,7 @@ pub struct SearchContext<'ctx> {
     pub phrase_interner: DedupInterner<Phrase>,
     pub term_interner: Interner<QueryTerm>,
     pub phrase_docids: PhraseDocIdsCache,
+    pub synonym_cache: SynonymCache,
     pub restricted_fids: Option<RestrictedFids>,
     pub prefix_search: PrefixSearch,
     pub vector_store_stats: Option<VectorStoreStats>,
@@ -103,6 +110,7 @@ impl<'ctx> SearchContext<'ctx> {
             phrase_interner: <_>::default(),
             term_interner: <_>::default(),
             phrase_docids: <_>::default(),
+            synonym_cache: <_>::default(),
             restricted_fids: None,
             prefix_search,
             vector_store_stats: None,
@@ -113,6 +121,17 @@ impl<'ctx> SearchContext<'ctx> {
         self.prefix_search != PrefixSearch::Disabled
     }
 
+    /// Get synonyms with caching to avoid repeated database access
+    pub fn get_synonyms(&mut self) -> Result<&HashMap<Vec<String>, Vec<Vec<String>>>> {
+        match self.synonym_cache.cache {
+            Some(ref synonyms) => Ok(synonyms),
+            None => {
+                let synonyms = self.index.synonyms(self.txn)?;
+                Ok(self.synonym_cache.cache.insert(synonyms))
+            }
+        }
+    }
+
     pub fn attributes_to_search_on(
         &mut self,
         attributes_to_search_on: &'ctx [String],

crates/milli/src/search/new/query_term/compute_derivations.rs

@@ -214,7 +214,7 @@ pub fn partially_initialized_term_from_word(
     if is_prefix && use_prefix_db.is_none() {
         find_zero_typo_prefix_derivations(ctx, word_interned, &mut prefix_of)?;
     }
-    let synonyms = ctx.index.synonyms(ctx.txn)?;
+    let synonyms = ctx.get_synonyms()?;
     let mut synonym_word_count = 0;
     let synonyms = synonyms
         .get(&vec![word.to_owned()])

crates/milli/src/search/new/query_term/parse_query.rs

@@ -258,7 +258,7 @@ pub fn make_ngram(
         partially_initialized_term_from_word(ctx, &ngram_str, max_nbr_typos, is_prefix, true)?;
 
     // Now add the synonyms
-    let index_synonyms = ctx.index.synonyms(ctx.txn)?;
+    let index_synonyms = ctx.get_synonyms()?;
 
     term.zero_typo.synonyms.extend(
         index_synonyms.get(&words).cloned().unwrap_or_default().into_iter().map(|words| {

crates/milli/src/search/new/ranking_rules.rs

@@ -1,9 +1,11 @@
+use std::task::Poll;
+
 use roaring::RoaringBitmap;
 
 use super::logger::SearchLogger;
 use super::{QueryGraph, SearchContext};
 use crate::score_details::ScoreDetails;
-use crate::Result;
+use crate::{Result, TimeBudget};
 
 /// An internal trait implemented by only [`PlaceholderQuery`] and [`QueryGraph`]
 pub trait RankingRuleQueryTrait: Sized + Clone + 'static {}
@@ -28,12 +30,15 @@ pub trait RankingRule<'ctx, Query: RankingRuleQueryTrait> {
     /// buckets using [`next_bucket`](RankingRule::next_bucket).
     ///
     /// The given universe is the universe that will be given to [`next_bucket`](RankingRule::next_bucket).
+    ///
+    /// If this function may take a long time, it should check the `time_budget` and return early if exceeded.
     fn start_iteration(
         &mut self,
         ctx: &mut SearchContext<'ctx>,
         logger: &mut dyn SearchLogger<Query>,
         universe: &RoaringBitmap,
         query: &Query,
+        time_budget: &TimeBudget,
     ) -> Result<()>;
 
     /// Return the next bucket of this ranking rule.
@@ -43,13 +48,31 @@ pub trait RankingRule<'ctx, Query: RankingRuleQueryTrait> {
     /// The universe given as argument is either:
     /// - a subset of the universe given to the previous call to [`next_bucket`](RankingRule::next_bucket); OR
     /// - the universe given to [`start_iteration`](RankingRule::start_iteration)
+    ///
+    /// If this function may take a long time, it should check the `time_budget` and return early if exceeded.
     fn next_bucket(
         &mut self,
         ctx: &mut SearchContext<'ctx>,
         logger: &mut dyn SearchLogger<Query>,
         universe: &RoaringBitmap,
+        time_budget: &TimeBudget,
     ) -> Result<Option<RankingRuleOutput<Query>>>;
 
+    /// Return the next bucket of this ranking rule, if doing so can be done without blocking
+    ///
+    /// Even if the time budget is exceeded, when getting the next bucket is a fast operation, this should return `true`
+    /// to allow Meilisearch to collect the results.
+    ///
+    /// Default implementation conservatively returns that it would block.
+    fn non_blocking_next_bucket(
+        &mut self,
+        _ctx: &mut SearchContext<'ctx>,
+        _logger: &mut dyn SearchLogger<Query>,
+        _universe: &RoaringBitmap,
+    ) -> Result<Poll<RankingRuleOutput<Query>>> {
+        Ok(Poll::Pending)
+    }
+
     /// Finish iterating over the buckets, which yields control to the parent ranking rule
     /// The next call to this ranking rule, if any, will be [`start_iteration`](RankingRule::start_iteration).
     fn end_iteration(

crates/milli/src/search/new/sort.rs

@@ -7,7 +7,7 @@ use crate::heed_codec::facet::{FacetGroupKeyCodec, OrderedF64Codec};
 use crate::heed_codec::{BytesRefCodec, StrRefCodec};
 use crate::score_details::{self, ScoreDetails};
 use crate::search::facet::{ascending_facet_sort, descending_facet_sort};
-use crate::{FieldId, Index, Result};
+use crate::{FieldId, Index, Result, TimeBudget};
 
 pub trait RankingRuleOutputIter<'ctx, Query> {
     fn next_bucket(&mut self) -> Result<Option<RankingRuleOutput<Query>>>;
@@ -96,6 +96,7 @@ impl<'ctx, Query: RankingRuleQueryTrait> RankingRule<'ctx, Query> for Sort<'ctx,
         _logger: &mut dyn SearchLogger<Query>,
         parent_candidates: &RoaringBitmap,
         parent_query: &Query,
+        _time_budget: &TimeBudget,
     ) -> Result<()> {
         let iter: RankingRuleOutputIterWrapper<'ctx, Query> = match self.field_id {
             Some(field_id) => {
@@ -194,6 +195,7 @@ impl<'ctx, Query: RankingRuleQueryTrait> RankingRule<'ctx, Query> for Sort<'ctx,
         _ctx: &mut SearchContext<'ctx>,
         _logger: &mut dyn SearchLogger<Query>,
         universe: &RoaringBitmap,
+        _time_budget: &TimeBudget,
     ) -> Result<Option<RankingRuleOutput<Query>>> {
         let iter = self.iter.as_mut().unwrap();
         if let Some(mut bucket) = iter.next_bucket()? {

crates/milli/src/search/new/tests/cutoff.rs

@@ -3,12 +3,17 @@
 //! 2. A test that ensure the filters are affectively applied even with a cutoff of 0
 //! 3. A test that ensure the cutoff works well with the ranking scores
 
+use std::collections::BTreeMap;
+use std::sync::Arc;
 use std::time::Duration;
 
 use meili_snap::snapshot;
 
 use crate::index::tests::TempIndex;
 use crate::score_details::{ScoreDetails, ScoringStrategy};
+use crate::update::Setting;
+use crate::vector::settings::EmbeddingSettings;
+use crate::vector::{Embedder, EmbedderOptions};
 use crate::{Criterion, Filter, FilterableAttributesRule, Search, TimeBudget};
 
 fn create_index() -> TempIndex {
@@ -361,9 +366,8 @@ fn degraded_search_and_score_details() {
     ]
     "###);
 
-    // After SIX loop iteration. The words ranking rule gave us a new bucket.
-    // Since we reached the limit we were able to early exit without checking the typo ranking rule.
-    search.time_budget(TimeBudget::max().with_stop_after(6));
+    // After FIVE loop iterations. The words ranking rule gave us a new bucket.
+    search.time_budget(TimeBudget::max().with_stop_after(5));
 
     let result = search.execute().unwrap();
     snapshot!(format!("IDs: {:?}\nScores: {}\nScore Details:\n{:#?}", result.documents_ids, result.document_scores.iter().map(|scores| format!("{:.4} ", ScoreDetails::global_score(scores.iter()))).collect::<String>(), result.document_scores), @r###"
@@ -424,4 +428,399 @@ fn degraded_search_and_score_details() {
         ],
     ]
     "###);
+
+    // After SIX loop iterations.
+    // we finished
+    search.time_budget(TimeBudget::max().with_stop_after(6));
+
+    let result = search.execute().unwrap();
+    snapshot!(format!("IDs: {:?}\nScores: {}\nScore Details:\n{:#?}", result.documents_ids, result.document_scores.iter().map(|scores| format!("{:.4} ", ScoreDetails::global_score(scores.iter()))).collect::<String>(), result.document_scores), @r###"
+    IDs: [4, 1, 0, 3]
+    Scores: 1.0000 0.9167 0.8333 0.6667 
+    Score Details:
+    [
+        [
+            Words(
+                Words {
+                    matching_words: 3,
+                    max_matching_words: 3,
+                },
+            ),
+            Typo(
+                Typo {
+                    typo_count: 0,
+                    max_typo_count: 3,
+                },
+            ),
+        ],
+        [
+            Words(
+                Words {
+                    matching_words: 3,
+                    max_matching_words: 3,
+                },
+            ),
+            Typo(
+                Typo {
+                    typo_count: 1,
+                    max_typo_count: 3,
+                },
+            ),
+        ],
+        [
+            Words(
+                Words {
+                    matching_words: 3,
+                    max_matching_words: 3,
+                },
+            ),
+            Typo(
+                Typo {
+                    typo_count: 2,
+                    max_typo_count: 3,
+                },
+            ),
+        ],
+        [
+            Words(
+                Words {
+                    matching_words: 2,
+                    max_matching_words: 3,
+                },
+            ),
+            Typo(
+                Typo {
+                    typo_count: 0,
+                    max_typo_count: 2,
+                },
+            ),
+        ],
+    ]
+    "###);
+}
+
+#[test]
+fn degraded_search_and_score_details_vector() {
+    let index = create_index();
+
+    index
+        .add_documents(documents!([
+            {
+                "id": 4,
+                "text": "hella puppo kefir",
+                "_vectors": {
+                    "default": [0.1, 0.1]
+                }
+            },
+            {
+                "id": 3,
+                "text": "hella puppy kefir",
+                "_vectors": {
+                    "default": [-0.1, 0.1]
+                }
+            },
+            {
+                "id": 2,
+                "text": "hello",
+                "_vectors": {
+                    "default": [0.1, -0.1]
+                }
+            },
+            {
+                "id": 1,
+                "text": "hello puppy",
+                "_vectors": {
+                    "default": [-0.1, -0.1]
+                }
+            },
+            {
+                "id": 0,
+                "text": "hello puppy kefir",
+                "_vectors": {
+                    "default": null
+                }
+            },
+        ]))
+        .unwrap();
+
+    index
+        .update_settings(|settings| {
+            let mut embedders = BTreeMap::new();
+            embedders.insert(
+                "default".into(),
+                Setting::Set(EmbeddingSettings {
+                    source: Setting::Set(crate::vector::settings::EmbedderSource::UserProvided),
+                    dimensions: Setting::Set(2),
+                    ..Default::default()
+                }),
+            );
+            settings.set_embedder_settings(embedders);
+            settings.set_vector_store(crate::vector::VectorStoreBackend::Hannoy);
+        })
+        .unwrap();
+
+    let rtxn = index.read_txn().unwrap();
+    let mut search = Search::new(&rtxn, &index);
+
+    let embedder = Arc::new(
+        Embedder::new(
+            EmbedderOptions::UserProvided(crate::vector::embedder::manual::EmbedderOptions {
+                dimensions: 2,
+                distribution: None,
+            }),
+            0,
+        )
+        .unwrap(),
+    );
+
+    search.semantic("default".into(), embedder, false, Some(vec![1., -1.]), None);
+
+    search.limit(4);
+    search.scoring_strategy(ScoringStrategy::Detailed);
+    search.time_budget(TimeBudget::max());
+
+    let result = search.execute().unwrap();
+    snapshot!(format!("IDs: {:?}\nScores: {}\nScore Details:\n{:#?}", result.documents_ids, result.document_scores.iter().map(|scores| format!("{:.4} ", ScoreDetails::global_score(scores.iter()))).collect::<String>(), result.document_scores), @r###"
+    IDs: [2, 0, 3, 1]
+    Scores: 1.0000 0.5000 0.5000 0.0000 
+    Score Details:
+    [
+        [
+            Vector(
+                Vector {
+                    similarity: Some(
+                        1.0,
+                    ),
+                },
+            ),
+        ],
+        [
+            Vector(
+                Vector {
+                    similarity: Some(
+                        0.5,
+                    ),
+                },
+            ),
+        ],
+        [
+            Vector(
+                Vector {
+                    similarity: Some(
+                        0.5,
+                    ),
+                },
+            ),
+        ],
+        [
+            Vector(
+                Vector {
+                    similarity: Some(
+                        0.0,
+                    ),
+                },
+            ),
+        ],
+    ]
+    "###);
+
+    // Do ONE loop iteration. Not much can be deduced, almost everyone matched the words first bucket.
+    search.time_budget(TimeBudget::max().with_stop_after(1));
+
+    let result = search.execute().unwrap();
+    snapshot!(format!("IDs: {:?}\nScores: {}\nScore Details:\n{:#?}", result.documents_ids, result.document_scores.iter().map(|scores| format!("{:.4} ", ScoreDetails::global_score(scores.iter()))).collect::<String>(), result.document_scores), @r###"
+    IDs: [0, 1, 2, 3]
+    Scores: 0.5000 0.0000 0.0000 0.0000 
+    Score Details:
+    [
+        [
+            Vector(
+                Vector {
+                    similarity: Some(
+                        0.5,
+                    ),
+                },
+            ),
+        ],
+        [
+            Skipped,
+        ],
+        [
+            Skipped,
+        ],
+        [
+            Skipped,
+        ],
+    ]
+    "###);
+
+    search.time_budget(TimeBudget::max().with_stop_after(2));
+
+    let result = search.execute().unwrap();
+    snapshot!(format!("IDs: {:?}\nScores: {}\nScore Details:\n{:#?}", result.documents_ids, result.document_scores.iter().map(|scores| format!("{:.4} ", ScoreDetails::global_score(scores.iter()))).collect::<String>(), result.document_scores), @r###"
+    IDs: [0, 1, 2, 3]
+    Scores: 0.5000 0.0000 0.0000 0.0000 
+    Score Details:
+    [
+        [
+            Vector(
+                Vector {
+                    similarity: Some(
+                        0.5,
+                    ),
+                },
+            ),
+        ],
+        [
+            Vector(
+                Vector {
+                    similarity: Some(
+                        0.0,
+                    ),
+                },
+            ),
+        ],
+        [
+            Skipped,
+        ],
+        [
+            Skipped,
+        ],
+    ]
+    "###);
+
+    search.time_budget(TimeBudget::max().with_stop_after(3));
+
+    let result = search.execute().unwrap();
+    snapshot!(format!("IDs: {:?}\nScores: {}\nScore Details:\n{:#?}", result.documents_ids, result.document_scores.iter().map(|scores| format!("{:.4} ", ScoreDetails::global_score(scores.iter()))).collect::<String>(), result.document_scores), @r###"
+    IDs: [2, 0, 1, 3]
+    Scores: 1.0000 0.5000 0.0000 0.0000 
+    Score Details:
+    [
+        [
+            Vector(
+                Vector {
+                    similarity: Some(
+                        1.0,
+                    ),
+                },
+            ),
+        ],
+        [
+            Vector(
+                Vector {
+                    similarity: Some(
+                        0.5,
+                    ),
+                },
+            ),
+        ],
+        [
+            Vector(
+                Vector {
+                    similarity: Some(
+                        0.0,
+                    ),
+                },
+            ),
+        ],
+        [
+            Skipped,
+        ],
+    ]
+    "###);
+
+    search.time_budget(TimeBudget::max().with_stop_after(4));
+
+    let result = search.execute().unwrap();
+    snapshot!(format!("IDs: {:?}\nScores: {}\nScore Details:\n{:#?}", result.documents_ids, result.document_scores.iter().map(|scores| format!("{:.4} ", ScoreDetails::global_score(scores.iter()))).collect::<String>(), result.document_scores), @r###"
+    IDs: [2, 0, 3, 1]
+    Scores: 1.0000 0.5000 0.5000 0.0000 
+    Score Details:
+    [
+        [
+            Vector(
+                Vector {
+                    similarity: Some(
+                        1.0,
+                    ),
+                },
+            ),
+        ],
+        [
+            Vector(
+                Vector {
+                    similarity: Some(
+                        0.5,
+                    ),
+                },
+            ),
+        ],
+        [
+            Vector(
+                Vector {
+                    similarity: Some(
+                        0.5,
+                    ),
+                },
+            ),
+        ],
+        [
+            Vector(
+                Vector {
+                    similarity: Some(
+                        0.0,
+                    ),
+                },
+            ),
+        ],
+    ]
+    "###);
+
+    search.time_budget(TimeBudget::max().with_stop_after(5));
+
+    let result = search.execute().unwrap();
+    snapshot!(format!("IDs: {:?}\nScores: {}\nScore Details:\n{:#?}", result.documents_ids, result.document_scores.iter().map(|scores| format!("{:.4} ", ScoreDetails::global_score(scores.iter()))).collect::<String>(), result.document_scores), @r###"
+    IDs: [2, 0, 3, 1]
+    Scores: 1.0000 0.5000 0.5000 0.0000 
+    Score Details:
+    [
+        [
+            Vector(
+                Vector {
+                    similarity: Some(
+                        1.0,
+                    ),
+                },
+            ),
+        ],
+        [
+            Vector(
+                Vector {
+                    similarity: Some(
+                        0.5,
+                    ),
+                },
+            ),
+        ],
+        [
+            Vector(
+                Vector {
+                    similarity: Some(
+                        0.5,
+                    ),
+                },
+            ),
+        ],
+        [
+            Vector(
+                Vector {
+                    similarity: Some(
+                        0.0,
+                    ),
+                },
+            ),
+        ],
+    ]
+    "###);
 }

crates/milli/src/search/new/vector_sort.rs

@@ -1,4 +1,5 @@
 use std::iter::FromIterator;
+use std::task::Poll;
 use std::time::Instant;
 
 use roaring::RoaringBitmap;
@@ -7,7 +8,7 @@ use super::ranking_rules::{RankingRule, RankingRuleOutput, RankingRuleQueryTrait
 use super::VectorStoreStats;
 use crate::score_details::{self, ScoreDetails};
 use crate::vector::{DistributionShift, Embedder, VectorStore};
-use crate::{DocumentId, Result, SearchContext, SearchLogger};
+use crate::{DocumentId, Result, SearchContext, SearchLogger, TimeBudget};
 
 pub struct VectorSort<Q: RankingRuleQueryTrait> {
     query: Option<Q>,
@@ -52,6 +53,7 @@ impl<Q: RankingRuleQueryTrait> VectorSort<Q> {
         &mut self,
         ctx: &mut SearchContext<'_>,
         vector_candidates: &RoaringBitmap,
+        time_budget: &TimeBudget,
     ) -> Result<()> {
         let target = &self.target;
         let backend = ctx.index.get_vector_store(ctx.txn)?.unwrap_or_default();
@@ -59,7 +61,13 @@ impl<Q: RankingRuleQueryTrait> VectorSort<Q> {
         let before = Instant::now();
         let reader =
             VectorStore::new(backend, ctx.index.vector_store, self.embedder_index, self.quantized);
-        let results = reader.nns_by_vector(ctx.txn, target, self.limit, Some(vector_candidates))?;
+        let results = reader.nns_by_vector(
+            ctx.txn,
+            target,
+            self.limit,
+            Some(vector_candidates),
+            time_budget,
+        )?;
         self.cached_sorted_docids = results.into_iter();
         *ctx.vector_store_stats.get_or_insert_default() += VectorStoreStats {
             total_time: before.elapsed(),
@@ -69,6 +77,20 @@ impl<Q: RankingRuleQueryTrait> VectorSort<Q> {
 
         Ok(())
     }
+
+    fn next_result(&mut self, vector_candidates: &RoaringBitmap) -> Option<(DocumentId, f32)> {
+        for (docid, distance) in self.cached_sorted_docids.by_ref() {
+            if vector_candidates.contains(docid) {
+                let score = 1.0 - distance;
+                let score = self
+                    .distribution_shift
+                    .map(|distribution| distribution.shift(score))
+                    .unwrap_or(score);
+                return Some((docid, score));
+            }
+        }
+        None
+    }
 }
 
 impl<'ctx, Q: RankingRuleQueryTrait> RankingRule<'ctx, Q> for VectorSort<Q> {
@@ -83,12 +105,13 @@ impl<'ctx, Q: RankingRuleQueryTrait> RankingRule<'ctx, Q> for VectorSort<Q> {
         _logger: &mut dyn SearchLogger<Q>,
         universe: &RoaringBitmap,
         query: &Q,
+        time_budget: &TimeBudget,
     ) -> Result<()> {
         assert!(self.query.is_none());
 
         self.query = Some(query.clone());
         let vector_candidates = &self.vector_candidates & universe;
-        self.fill_buffer(ctx, &vector_candidates)?;
+        self.fill_buffer(ctx, &vector_candidates, time_budget)?;
         Ok(())
     }
 
@@ -99,6 +122,7 @@ impl<'ctx, Q: RankingRuleQueryTrait> RankingRule<'ctx, Q> for VectorSort<Q> {
         ctx: &mut SearchContext<'ctx>,
         _logger: &mut dyn SearchLogger<Q>,
         universe: &RoaringBitmap,
+        time_budget: &TimeBudget,
     ) -> Result<Option<RankingRuleOutput<Q>>> {
         let query = self.query.as_ref().unwrap().clone();
         let vector_candidates = &self.vector_candidates & universe;
@@ -111,24 +135,17 @@ impl<'ctx, Q: RankingRuleQueryTrait> RankingRule<'ctx, Q> for VectorSort<Q> {
             }));
         }
 
-        for (docid, distance) in self.cached_sorted_docids.by_ref() {
-            if vector_candidates.contains(docid) {
-                let score = 1.0 - distance;
-                let score = self
-                    .distribution_shift
-                    .map(|distribution| distribution.shift(score))
-                    .unwrap_or(score);
-                return Ok(Some(RankingRuleOutput {
-                    query,
-                    candidates: RoaringBitmap::from_iter([docid]),
-                    score: ScoreDetails::Vector(score_details::Vector { similarity: Some(score) }),
-                }));
-            }
+        if let Some((docid, score)) = self.next_result(&vector_candidates) {
+            return Ok(Some(RankingRuleOutput {
+                query,
+                candidates: RoaringBitmap::from_iter([docid]),
+                score: ScoreDetails::Vector(score_details::Vector { similarity: Some(score) }),
+            }));
         }
 
         // if we got out of this loop it means we've exhausted our cache.
         // we need to refill it and run the function again.
-        self.fill_buffer(ctx, &vector_candidates)?;
+        self.fill_buffer(ctx, &vector_candidates, time_budget)?;
 
         // we tried filling the buffer, but it remained empty 😢
         // it means we don't actually have any document remaining in the universe with a vector.
@@ -141,11 +158,39 @@ impl<'ctx, Q: RankingRuleQueryTrait> RankingRule<'ctx, Q> for VectorSort<Q> {
             }));
         }
 
-        self.next_bucket(ctx, _logger, universe)
+        self.next_bucket(ctx, _logger, universe, time_budget)
     }
 
     #[tracing::instrument(level = "trace", skip_all, target = "search::vector_sort")]
     fn end_iteration(&mut self, _ctx: &mut SearchContext<'ctx>, _logger: &mut dyn SearchLogger<Q>) {
         self.query = None;
     }
+
+    fn non_blocking_next_bucket(
+        &mut self,
+        _ctx: &mut SearchContext<'ctx>,
+        _logger: &mut dyn SearchLogger<Q>,
+        universe: &RoaringBitmap,
+    ) -> Result<Poll<RankingRuleOutput<Q>>> {
+        let query = self.query.as_ref().unwrap().clone();
+        let vector_candidates = &self.vector_candidates & universe;
+
+        if vector_candidates.is_empty() {
+            return Ok(Poll::Ready(RankingRuleOutput {
+                query,
+                candidates: universe.clone(),
+                score: ScoreDetails::Vector(score_details::Vector { similarity: None }),
+            }));
+        }
+
+        if let Some((docid, score)) = self.next_result(&vector_candidates) {
+            Ok(Poll::Ready(RankingRuleOutput {
+                query,
+                candidates: RoaringBitmap::from_iter([docid]),
+                score: ScoreDetails::Vector(score_details::Vector { similarity: Some(score) }),
+            }))
+        } else {
+            Ok(Poll::Pending)
+        }
+    }
 }

crates/milli/src/update/index_documents/helpers/grenad_helpers.rs

@@ -101,6 +101,8 @@ pub struct GrenadParameters {
     pub chunk_compression_level: Option<u32>,
     pub max_memory: Option<usize>,
     pub max_nb_chunks: Option<usize>,
+    pub experimental_no_edition_2024_for_prefix_post_processing: bool,
+    pub experimental_no_edition_2024_for_facet_post_processing: bool,
 }
 
 impl Default for GrenadParameters {
@@ -110,6 +112,8 @@ impl Default for GrenadParameters {
             chunk_compression_level: None,
             max_memory: None,
             max_nb_chunks: None,
+            experimental_no_edition_2024_for_prefix_post_processing: false,
+            experimental_no_edition_2024_for_facet_post_processing: false,
         }
     }
 }

crates/milli/src/update/index_documents/mod.rs

@@ -254,6 +254,12 @@ where
             chunk_compression_level: self.indexer_config.chunk_compression_level,
             max_memory: self.indexer_config.max_memory,
             max_nb_chunks: self.indexer_config.max_nb_chunks, // default value, may be chosen.
+            experimental_no_edition_2024_for_prefix_post_processing: self
+                .indexer_config
+                .experimental_no_edition_2024_for_prefix_post_processing,
+            experimental_no_edition_2024_for_facet_post_processing: self
+                .indexer_config
+                .experimental_no_edition_2024_for_facet_post_processing,
         };
         let documents_chunk_size = match self.indexer_config.documents_chunk_size {
             Some(chunk_size) => chunk_size,

crates/milli/src/update/index_documents/transform.rs

@@ -983,6 +983,12 @@ impl<'a, 'i> Transform<'a, 'i> {
             chunk_compression_level: self.indexer_settings.chunk_compression_level,
             max_memory: self.indexer_settings.max_memory,
             max_nb_chunks: self.indexer_settings.max_nb_chunks, // default value, may be chosen.
+            experimental_no_edition_2024_for_prefix_post_processing: self
+                .indexer_settings
+                .experimental_no_edition_2024_for_prefix_post_processing,
+            experimental_no_edition_2024_for_facet_post_processing: self
+                .indexer_settings
+                .experimental_no_edition_2024_for_facet_post_processing,
         };
 
         // Once we have written all the documents, we merge everything into a Reader.

crates/milli/src/update/indexer_config.rs

@@ -18,6 +18,8 @@ pub struct IndexerConfig {
     pub skip_index_budget: bool,
     pub experimental_no_edition_2024_for_settings: bool,
     pub experimental_no_edition_2024_for_dumps: bool,
+    pub experimental_no_edition_2024_for_prefix_post_processing: bool,
+    pub experimental_no_edition_2024_for_facet_post_processing: bool,
 }
 
 impl IndexerConfig {
@@ -27,6 +29,10 @@ impl IndexerConfig {
             chunk_compression_level: self.chunk_compression_level,
             max_memory: self.max_memory,
             max_nb_chunks: self.max_nb_chunks,
+            experimental_no_edition_2024_for_prefix_post_processing: self
+                .experimental_no_edition_2024_for_prefix_post_processing,
+            experimental_no_edition_2024_for_facet_post_processing: self
+                .experimental_no_edition_2024_for_facet_post_processing,
         }
     }
 }
@@ -68,6 +74,8 @@ impl Default for IndexerConfig {
             skip_index_budget: false,
             experimental_no_edition_2024_for_settings: false,
             experimental_no_edition_2024_for_dumps: false,
+            experimental_no_edition_2024_for_prefix_post_processing: false,
+            experimental_no_edition_2024_for_facet_post_processing: false,
         }
     }
 }

crates/milli/src/update/new/indexer/mod.rs

@@ -180,12 +180,15 @@ where
         })
         .unwrap()?;
 
-        post_processing::post_process(
-            indexing_context,
-            wtxn,
-            global_fields_ids_map,
-            facet_field_ids_delta,
-        )?;
+        pool.install(|| {
+            post_processing::post_process(
+                indexing_context,
+                wtxn,
+                global_fields_ids_map,
+                facet_field_ids_delta,
+            )
+        })
+        .unwrap()?;
 
         indexing_context.progress.update_progress(IndexingStep::BuildingGeoJson);
         index.cellulite.build(

crates/milli/src/update/new/indexer/post_processing/facet_bulk.rs

@@ -0,0 +1,164 @@
+use std::fs::File;
+use std::io::BufReader;
+use std::{iter, mem};
+
+use grenad::CompressionType;
+use heed::types::{Bytes, LazyDecode};
+use heed::{Database, RwTxn};
+use rayon::prelude::*;
+use roaring::MultiOps;
+use tempfile::tempfile;
+
+use crate::facet::FacetType;
+use crate::heed_codec::facet::{FacetGroupKey, FacetGroupKeyCodec, FacetGroupValueCodec};
+use crate::heed_codec::BytesRefCodec;
+use crate::update::facet::{FACET_GROUP_SIZE, FACET_MIN_LEVEL_SIZE};
+use crate::update::{create_writer, writer_into_reader};
+use crate::{CboRoaringBitmapCodec, FieldId, Index};
+
+/// Generate the facet level based on the level 0.
+///
+/// The function will generate all the group levels from
+/// the group 1 to the level n until the number of group
+/// is smaller than the minimum required size.
+pub fn generate_facet_levels(
+    index: &Index,
+    wtxn: &mut RwTxn,
+    field_id: FieldId,
+    facet_type: FacetType,
+) -> crate::Result<()> {
+    let db = match facet_type {
+        FacetType::String => index
+            .facet_id_string_docids
+            .remap_key_type::<FacetGroupKeyCodec<BytesRefCodec>>()
+            .lazily_decode_data(),
+        FacetType::Number => index
+            .facet_id_f64_docids
+            .remap_key_type::<FacetGroupKeyCodec<BytesRefCodec>>()
+            .lazily_decode_data(),
+    };
+
+    clear_levels(db, wtxn, field_id)?;
+
+    let mut base_level: u8 = 0;
+    // That's a do-while loop
+    while {
+        let mut level_size = 0;
+        let level = base_level.checked_add(1).unwrap();
+        for reader in compute_level(index, wtxn, db, field_id, base_level)? {
+            let mut cursor = reader.into_cursor()?;
+            while let Some((left_bound, facet_group_value)) = cursor.move_on_next()? {
+                level_size += 1;
+                let key = FacetGroupKey { field_id, level, left_bound };
+                debug_assert!(
+                    db.get(wtxn, &key).transpose().is_none(),
+                    "entry must not be there and must have already been deleted: {key:?}"
+                );
+                db.remap_data_type::<Bytes>().put(wtxn, &key, facet_group_value)?;
+            }
+        }
+
+        base_level = level;
+
+        // If the next level will have the minimum required groups, continue.
+        (level_size / FACET_GROUP_SIZE as usize) >= FACET_MIN_LEVEL_SIZE as usize
+    } {}
+
+    Ok(())
+}
+
+/// Compute the groups of facets from the provided base level
+/// and write the content into different grenad files.
+fn compute_level(
+    index: &Index,
+    wtxn: &heed::RwTxn,
+    db: Database<FacetGroupKeyCodec<BytesRefCodec>, LazyDecode<FacetGroupValueCodec>>,
+    field_id: FieldId,
+    base_level: u8,
+) -> Result<Vec<grenad::Reader<BufReader<File>>>, crate::Error> {
+    let thread_count = rayon::current_num_threads();
+    let rtxns = iter::repeat_with(|| index.env.nested_read_txn(wtxn))
+        .take(thread_count)
+        .collect::<heed::Result<Vec<_>>>()?;
+
+    let range = {
+        // Based on the first possible value for the base level up to
+        // the first possible value for the next level *excluded*.
+        let left = FacetGroupKey::<&[u8]> { field_id, level: base_level, left_bound: &[] };
+        let right = FacetGroupKey::<&[u8]> {
+            field_id,
+            level: base_level.checked_add(1).unwrap(),
+            left_bound: &[],
+        };
+        left..right
+    };
+
+    rtxns
+        .into_par_iter()
+        .enumerate()
+        .map(|(thread_id, rtxn)| {
+            let mut writer = tempfile().map(|f| create_writer(CompressionType::None, None, f))?;
+
+            let mut left_bound = None;
+            let mut group_docids = Vec::new();
+            let mut ser_buffer = Vec::new();
+            for (i, result) in db.range(&rtxn, &range)?.enumerate() {
+                let (key, lazy_value) = result?;
+
+                let start_of_group = i % FACET_GROUP_SIZE as usize == 0;
+                let group_index = i / FACET_GROUP_SIZE as usize;
+                let group_for_thread = group_index % thread_count == thread_id;
+
+                if group_for_thread {
+                    if start_of_group {
+                        if let Some(left_bound) = left_bound.take() {
+                            // We store the bitmaps in a Vec this way we can use
+                            // the MultiOps operations that tends to be more efficient
+                            // for unions. The Vec is empty after the operation.
+                            //
+                            // We also don't forget to store the group size corresponding
+                            // to the number of entries merged in this group.
+                            ser_buffer.clear();
+                            let group_len: u8 = group_docids.len().try_into().unwrap();
+                            ser_buffer.push(group_len);
+                            let group_docids = mem::take(&mut group_docids);
+                            let docids = group_docids.into_iter().union();
+                            CboRoaringBitmapCodec::serialize_into_vec(&docids, &mut ser_buffer);
+                            writer.insert(left_bound, &ser_buffer)?;
+                        }
+                        left_bound = Some(key.left_bound);
+                    }
+
+                    // Lazily decode the bitmaps we are interested in.
+                    let value = lazy_value.decode().map_err(heed::Error::Decoding)?;
+                    group_docids.push(value.bitmap);
+                }
+            }
+
+            if let Some(left_bound) = left_bound.take() {
+                ser_buffer.clear();
+                // We don't forget to store the group size corresponding
+                // to the number of entries merged in this group.
+                let group_len: u8 = group_docids.len().try_into().unwrap();
+                ser_buffer.push(group_len);
+                let group_docids = group_docids.into_iter().union();
+                CboRoaringBitmapCodec::serialize_into_vec(&group_docids, &mut ser_buffer);
+                writer.insert(left_bound, &ser_buffer)?;
+            }
+
+            writer_into_reader(writer)
+        })
+        .collect()
+}
+
+/// Clears all the levels and only keeps the level 0 of the specified field id.
+fn clear_levels(
+    db: Database<FacetGroupKeyCodec<BytesRefCodec>, LazyDecode<FacetGroupValueCodec>>,
+    wtxn: &mut RwTxn<'_>,
+    field_id: FieldId,
+) -> heed::Result<()> {
+    let left = FacetGroupKey::<&[u8]> { field_id, level: 1, left_bound: &[] };
+    let right = FacetGroupKey::<&[u8]> { field_id, level: u8::MAX, left_bound: &[] };
+    let range = left..=right;
+    db.delete_range(wtxn, &range).map(drop)
+}