change the version of arroy

5376: Support dumpless upgrade for all v1.13 patches r=Kerollmops a=dureuill

# Pull Request

## Related issue
Fixes #5373 

## What does this PR do?
- ensure v1.13 versions are known to the index scheduler upgrading code
- Forbid opening a db of v1.13.x from v1.13.y
- Keep old stat format to make sure the number of documents is available in stats during dumpless upgrade


Co-authored-by: Kerollmops <Kerollmops@users.noreply.github.com>
Co-authored-by: Louis Dureuil <louis@meilisearch.com>

Cargo.lock

@@ -394,8 +394,7 @@ checksum = "96d30a06541fbafbc7f82ed10c06164cfbd2c401138f6addd8404629c4b16711"
 [[package]]
 name = "arroy"
 version = "0.5.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "dfc5f272f38fa063bbff0a7ab5219404e221493de005e2b4078c62d626ef567e"
+source = "git+https://github.com/meilisearch/arroy/?tag=DO-NOT-DELETE-upgrade-v04-to-v05#053807bf38dc079f25b003f19fc30fbf3613f6e7"
 dependencies = [
  "bytemuck",
  "byteorder",
@@ -414,7 +413,7 @@ dependencies = [
 [[package]]
 name = "arroy"
 version = "0.5.0"
-source = "git+https://github.com/meilisearch/arroy/?tag=DO-NOT-DELETE-upgrade-v04-to-v05#053807bf38dc079f25b003f19fc30fbf3613f6e7"
+source = "git+https://github.com/meilisearch/arroy?rev=55fb0e8006f4f00ad3197b06bda348133f6ffffa#55fb0e8006f4f00ad3197b06bda348133f6ffffa"
 dependencies = [
  "bytemuck",
  "byteorder",
@@ -427,7 +426,7 @@ dependencies = [
  "rayon",
  "roaring",
  "tempfile",
- "thiserror 1.0.69",
+ "thiserror 2.0.9",
 ]
 
 [[package]]
@@ -503,7 +502,7 @@ source = "git+https://github.com/meilisearch/bbqueue#cbb87cc707b5af415ef203bdaf2
 
 [[package]]
 name = "benchmarks"
-version = "1.13.0"
+version = "1.13.2"
 dependencies = [
  "anyhow",
  "bumpalo",
@@ -694,7 +693,7 @@ dependencies = [
 
 [[package]]
 name = "build-info"
-version = "1.13.0"
+version = "1.13.2"
 dependencies = [
  "anyhow",
  "time",
@@ -1671,7 +1670,7 @@ dependencies = [
 
 [[package]]
 name = "dump"
-version = "1.13.0"
+version = "1.13.2"
 dependencies = [
  "anyhow",
  "big_s",
@@ -1873,7 +1872,7 @@ checksum = "37909eebbb50d72f9059c3b6d82c0463f2ff062c9e95845c43a6c9c0355411be"
 
 [[package]]
 name = "file-store"
-version = "1.13.0"
+version = "1.13.2"
 dependencies = [
  "tempfile",
  "thiserror 2.0.9",
@@ -1895,7 +1894,7 @@ dependencies = [
 
 [[package]]
 name = "filter-parser"
-version = "1.13.0"
+version = "1.13.2"
 dependencies = [
  "insta",
  "nom",
@@ -1915,7 +1914,7 @@ dependencies = [
 
 [[package]]
 name = "flatten-serde-json"
-version = "1.13.0"
+version = "1.13.2"
 dependencies = [
  "criterion",
  "serde_json",
@@ -2054,7 +2053,7 @@ dependencies = [
 
 [[package]]
 name = "fuzzers"
-version = "1.13.0"
+version = "1.13.2"
 dependencies = [
  "arbitrary",
  "bumpalo",
@@ -2743,10 +2742,10 @@ checksum = "206ca75c9c03ba3d4ace2460e57b189f39f43de612c2f85836e65c929701bb2d"
 
 [[package]]
 name = "index-scheduler"
-version = "1.13.0"
+version = "1.13.2"
 dependencies = [
  "anyhow",
- "arroy 0.5.0 (registry+https://github.com/rust-lang/crates.io-index)",
+ "arroy 0.5.0 (git+https://github.com/meilisearch/arroy?rev=55fb0e8006f4f00ad3197b06bda348133f6ffffa)",
  "big_s",
  "bincode",
  "bumpalo",
@@ -2950,7 +2949,7 @@ dependencies = [
 
 [[package]]
 name = "json-depth-checker"
-version = "1.13.0"
+version = "1.13.2"
 dependencies = [
  "criterion",
  "serde_json",
@@ -3513,9 +3512,9 @@ checksum = "9374ef4228402d4b7e403e5838cb880d9ee663314b0a900d5a6aabf0c213552e"
 
 [[package]]
 name = "log"
-version = "0.4.21"
+version = "0.4.26"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "90ed8c1e510134f979dbc4f070f87d4313098b704861a105fe34231c70a3901c"
+checksum = "30bde2b3dc3671ae49d8e2e9f044c7c005836e7a023ee57cffa25ab82764bb9e"
 
 [[package]]
 name = "lzma-rs"
@@ -3569,7 +3568,7 @@ checksum = "490cc448043f947bae3cbee9c203358d62dbee0db12107a74be5c30ccfd09771"
 
 [[package]]
 name = "meili-snap"
-version = "1.13.0"
+version = "1.13.2"
 dependencies = [
  "insta",
  "md5",
@@ -3578,7 +3577,7 @@ dependencies = [
 
 [[package]]
 name = "meilisearch"
-version = "1.13.0"
+version = "1.13.2"
 dependencies = [
  "actix-cors",
  "actix-http",
@@ -3670,7 +3669,7 @@ dependencies = [
 
 [[package]]
 name = "meilisearch-auth"
-version = "1.13.0"
+version = "1.13.2"
 dependencies = [
  "base64 0.22.1",
  "enum-iterator",
@@ -3689,7 +3688,7 @@ dependencies = [
 
 [[package]]
 name = "meilisearch-types"
-version = "1.13.0"
+version = "1.13.2"
 dependencies = [
  "actix-web",
  "anyhow",
@@ -3723,7 +3722,7 @@ dependencies = [
 
 [[package]]
 name = "meilitool"
-version = "1.13.0"
+version = "1.13.2"
 dependencies = [
  "anyhow",
  "arroy 0.5.0 (git+https://github.com/meilisearch/arroy/?tag=DO-NOT-DELETE-upgrade-v04-to-v05)",
@@ -3758,10 +3757,10 @@ dependencies = [
 
 [[package]]
 name = "milli"
-version = "1.13.0"
+version = "1.13.2"
 dependencies = [
  "allocator-api2",
- "arroy 0.5.0 (registry+https://github.com/rust-lang/crates.io-index)",
+ "arroy 0.5.0 (git+https://github.com/meilisearch/arroy?rev=55fb0e8006f4f00ad3197b06bda348133f6ffffa)",
  "bbqueue",
  "big_s",
  "bimap",
@@ -4270,7 +4269,7 @@ checksum = "e3148f5046208a5d56bcfc03053e3ca6334e51da8dfb19b6cdc8b306fae3283e"
 
 [[package]]
 name = "permissive-json-pointer"
-version = "1.13.0"
+version = "1.13.2"
 dependencies = [
  "big_s",
  "serde_json",
@@ -6847,7 +6846,7 @@ dependencies = [
 
 [[package]]
 name = "xtask"
-version = "1.13.0"
+version = "1.13.2"
 dependencies = [
  "anyhow",
  "build-info",

Cargo.toml

@@ -22,7 +22,7 @@ members = [
 ]
 
 [workspace.package]
-version = "1.13.0"
+version = "1.13.2"
 authors = [
     "Quentin de Quelen <quentin@dequelen.me>",
     "Clément Renault <clement@meilisearch.com>",

crates/index-scheduler/Cargo.toml

@@ -44,7 +44,7 @@ ureq = "2.12.1"
 uuid = { version = "1.11.0", features = ["serde", "v4"] }
 
 [dev-dependencies]
-arroy = "0.5.0"
+arroy = { git = "https://github.com/meilisearch/arroy", rev = "55fb0e8006f4f00ad3197b06bda348133f6ffffa" }
 big_s = "1.0.2"
 crossbeam-channel = "0.5.14"
 # fixed version due to format breakages in v1.40

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

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

crates/index-scheduler/src/insta_snapshot.rs

@@ -365,7 +365,8 @@ pub fn snapshot_index_mapper(rtxn: &RoTxn, mapper: &IndexMapper) -> String {
         let stats = mapper.stats_of(rtxn, &name).unwrap();
         s.push_str(&format!(
             "{name}: {{ number_of_documents: {}, field_distribution: {:?} }}\n",
-            stats.number_of_documents, stats.field_distribution
+            stats.documents_database_stats.number_of_entries(),
+            stats.field_distribution
         ));
     }
 

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

@@ -1,13 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test_failure.rs
-snapshot_kind: text
 ---
 ### Autobatching Enabled = true
 ### Processing batch None:
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, batch_uid: 0, status: succeeded, details: { from: (1, 12, 0), to: (1, 13, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
+0 {uid: 0, batch_uid: 0, status: succeeded, details: { from: (1, 12, 0), to: (1, 13, 2) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
 1 {uid: 1, batch_uid: 1, status: succeeded, details: { primary_key: Some("mouse") }, kind: IndexCreation { index_uid: "catto", primary_key: Some("mouse") }}
 2 {uid: 2, batch_uid: 2, status: succeeded, details: { primary_key: Some("bone") }, kind: IndexCreation { index_uid: "doggo", primary_key: Some("bone") }}
 3 {uid: 3, batch_uid: 3, status: failed, error: ResponseError { code: 200, message: "Index `doggo` already exists.", error_code: "index_already_exists", error_type: "invalid_request", error_link: "https://docs.meilisearch.com/errors#index_already_exists" }, details: { primary_key: Some("bone") }, kind: IndexCreation { index_uid: "doggo", primary_key: Some("bone") }}
@@ -58,7 +57,7 @@ girafo: { number_of_documents: 0, field_distribution: {} }
 [timestamp] [4,]
 ----------------------------------------------------------------------
 ### All Batches:
-0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.13.0"}, stats: {"totalNbTasks":1,"status":{"succeeded":1},"types":{"upgradeDatabase":1},"indexUids":{}}, }
+0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.13.2"}, stats: {"totalNbTasks":1,"status":{"succeeded":1},"types":{"upgradeDatabase":1},"indexUids":{}}, }
 1 {uid: 1, details: {"primaryKey":"mouse"}, stats: {"totalNbTasks":1,"status":{"succeeded":1},"types":{"indexCreation":1},"indexUids":{"catto":1}}, }
 2 {uid: 2, details: {"primaryKey":"bone"}, stats: {"totalNbTasks":1,"status":{"succeeded":1},"types":{"indexCreation":1},"indexUids":{"doggo":1}}, }
 3 {uid: 3, details: {"primaryKey":"bone"}, stats: {"totalNbTasks":1,"status":{"failed":1},"types":{"indexCreation":1},"indexUids":{"doggo":1}}, }

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

@@ -1,13 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test_failure.rs
-snapshot_kind: text
 ---
 ### Autobatching Enabled = true
 ### Processing batch None:
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, status: enqueued, details: { from: (1, 12, 0), to: (1, 13, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
+0 {uid: 0, status: enqueued, details: { from: (1, 12, 0), to: (1, 13, 2) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
 ----------------------------------------------------------------------
 ### Status:
 enqueued [0,]

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

@@ -1,13 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test_failure.rs
-snapshot_kind: text
 ---
 ### Autobatching Enabled = true
 ### Processing batch None:
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, status: enqueued, details: { from: (1, 12, 0), to: (1, 13, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
+0 {uid: 0, status: enqueued, details: { from: (1, 12, 0), to: (1, 13, 2) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
 1 {uid: 1, status: enqueued, details: { primary_key: Some("mouse") }, kind: IndexCreation { index_uid: "catto", primary_key: Some("mouse") }}
 ----------------------------------------------------------------------
 ### Status:

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

@@ -1,13 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test_failure.rs
-snapshot_kind: text
 ---
 ### Autobatching Enabled = true
 ### Processing batch None:
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, batch_uid: 0, status: failed, error: ResponseError { code: 200, message: "Planned failure for tests.", error_code: "internal", error_type: "internal", error_link: "https://docs.meilisearch.com/errors#internal" }, details: { from: (1, 12, 0), to: (1, 13, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
+0 {uid: 0, batch_uid: 0, status: failed, error: ResponseError { code: 200, message: "Planned failure for tests.", error_code: "internal", error_type: "internal", error_link: "https://docs.meilisearch.com/errors#internal" }, details: { from: (1, 12, 0), to: (1, 13, 2) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
 1 {uid: 1, status: enqueued, details: { primary_key: Some("mouse") }, kind: IndexCreation { index_uid: "catto", primary_key: Some("mouse") }}
 ----------------------------------------------------------------------
 ### Status:
@@ -38,7 +37,7 @@ catto [1,]
 [timestamp] [0,]
 ----------------------------------------------------------------------
 ### All Batches:
-0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.13.0"}, stats: {"totalNbTasks":1,"status":{"failed":1},"types":{"upgradeDatabase":1},"indexUids":{}}, }
+0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.13.2"}, stats: {"totalNbTasks":1,"status":{"failed":1},"types":{"upgradeDatabase":1},"indexUids":{}}, }
 ----------------------------------------------------------------------
 ### Batch to tasks mapping:
 0 [0,]

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

@@ -1,13 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test_failure.rs
-snapshot_kind: text
 ---
 ### Autobatching Enabled = true
 ### Processing batch None:
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, batch_uid: 0, status: failed, error: ResponseError { code: 200, message: "Planned failure for tests.", error_code: "internal", error_type: "internal", error_link: "https://docs.meilisearch.com/errors#internal" }, details: { from: (1, 12, 0), to: (1, 13, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
+0 {uid: 0, batch_uid: 0, status: failed, error: ResponseError { code: 200, message: "Planned failure for tests.", error_code: "internal", error_type: "internal", error_link: "https://docs.meilisearch.com/errors#internal" }, details: { from: (1, 12, 0), to: (1, 13, 2) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
 1 {uid: 1, status: enqueued, details: { primary_key: Some("mouse") }, kind: IndexCreation { index_uid: "catto", primary_key: Some("mouse") }}
 2 {uid: 2, status: enqueued, details: { primary_key: Some("bone") }, kind: IndexCreation { index_uid: "doggo", primary_key: Some("bone") }}
 ----------------------------------------------------------------------
@@ -41,7 +40,7 @@ doggo [2,]
 [timestamp] [0,]
 ----------------------------------------------------------------------
 ### All Batches:
-0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.13.0"}, stats: {"totalNbTasks":1,"status":{"failed":1},"types":{"upgradeDatabase":1},"indexUids":{}}, }
+0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.13.2"}, stats: {"totalNbTasks":1,"status":{"failed":1},"types":{"upgradeDatabase":1},"indexUids":{}}, }
 ----------------------------------------------------------------------
 ### Batch to tasks mapping:
 0 [0,]

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

@@ -1,13 +1,12 @@
 ---
 source: crates/index-scheduler/src/scheduler/test_failure.rs
-snapshot_kind: text
 ---
 ### Autobatching Enabled = true
 ### Processing batch None:
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, batch_uid: 0, status: succeeded, details: { from: (1, 12, 0), to: (1, 13, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
+0 {uid: 0, batch_uid: 0, status: succeeded, details: { from: (1, 12, 0), to: (1, 13, 2) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
 1 {uid: 1, status: enqueued, details: { primary_key: Some("mouse") }, kind: IndexCreation { index_uid: "catto", primary_key: Some("mouse") }}
 2 {uid: 2, status: enqueued, details: { primary_key: Some("bone") }, kind: IndexCreation { index_uid: "doggo", primary_key: Some("bone") }}
 3 {uid: 3, status: enqueued, details: { primary_key: Some("bone") }, kind: IndexCreation { index_uid: "doggo", primary_key: Some("bone") }}
@@ -44,7 +43,7 @@ doggo [2,3,]
 [timestamp] [0,]
 ----------------------------------------------------------------------
 ### All Batches:
-0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.13.0"}, stats: {"totalNbTasks":1,"status":{"succeeded":1},"types":{"upgradeDatabase":1},"indexUids":{}}, }
+0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.13.2"}, stats: {"totalNbTasks":1,"status":{"succeeded":1},"types":{"upgradeDatabase":1},"indexUids":{}}, }
 ----------------------------------------------------------------------
 ### Batch to tasks mapping:
 0 [0,]

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

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

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

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

crates/meilisearch/Cargo.toml

@@ -169,5 +169,5 @@ german = ["meilisearch-types/german"]
 turkish = ["meilisearch-types/turkish"]
 
 [package.metadata.mini-dashboard]
-assets-url = "https://github.com/meilisearch/mini-dashboard/releases/download/v0.2.16/build.zip"
-sha1 = "68f83438a114aabbe76bc9fe480071e741996662"
+assets-url = "https://github.com/meilisearch/mini-dashboard/releases/download/v0.2.17/build.zip"
+sha1 = "29e92ce25f306208a9c86f013279c736bdc1e034"

crates/meilisearch/src/lib.rs

@@ -364,7 +364,7 @@ fn check_version(
     let (bin_major, bin_minor, bin_patch) = binary_version;
     let (db_major, db_minor, db_patch) = get_version(&opt.db_path)?;
 
-    if db_major != bin_major || db_minor != bin_minor || db_patch > bin_patch {
+    if db_major != bin_major || db_minor != bin_minor || db_patch != bin_patch {
         if opt.experimental_dumpless_upgrade {
             update_version_file_for_dumpless_upgrade(
                 opt,

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

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

crates/meilisearch/src/routes/mod.rs

@@ -392,6 +392,9 @@ pub struct Stats {
                 "indexes": {
                     "movies": {
                         "numberOfDocuments": 10,
+                        "rawDocumentDbSize": 100,
+                        "maxDocumentSize": 16,
+                        "avgDocumentSize": 10,
                         "isIndexing": true,
                         "fieldDistribution": {
                             "genre": 10,

crates/meilisearch/tests/documents/add_documents.rs

@@ -1803,6 +1803,275 @@ async fn add_documents_with_geo_field() {
       "finishedAt": "[date]"
     }
     "###);
+
+    let (response, code) = index.get_all_documents(GetAllDocumentsOptions::default()).await;
+
+    snapshot!(code, @"200 OK");
+    snapshot!(json_string!(response, { ".duration" => "[duration]", ".enqueuedAt" => "[date]", ".startedAt" => "[date]", ".finishedAt" => "[date]" }),
+    @r###"
+    {
+      "results": [
+        {
+          "id": "1"
+        },
+        {
+          "id": "2",
+          "_geo": null
+        },
+        {
+          "id": "3",
+          "_geo": {
+            "lat": 1,
+            "lng": 1
+          }
+        },
+        {
+          "id": "4",
+          "_geo": {
+            "lat": "1",
+            "lng": "1"
+          }
+        }
+      ],
+      "offset": 0,
+      "limit": 20,
+      "total": 4
+    }
+    "###);
+
+    let (response, code) = index
+        .search_post(json!({"sort": ["_geoPoint(50.629973371633746,3.0569447399419567):desc"]}))
+        .await;
+    snapshot!(code, @"200 OK");
+    // we are expecting docs 4 and 3 first as they have geo
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]" }),
+    @r###"
+    {
+      "hits": [
+        {
+          "id": "4",
+          "_geo": {
+            "lat": "1",
+            "lng": "1"
+          },
+          "_geoDistance": 5522018
+        },
+        {
+          "id": "3",
+          "_geo": {
+            "lat": 1,
+            "lng": 1
+          },
+          "_geoDistance": 5522018
+        },
+        {
+          "id": "1"
+        },
+        {
+          "id": "2",
+          "_geo": null
+        }
+      ],
+      "query": "",
+      "processingTimeMs": "[time]",
+      "limit": 20,
+      "offset": 0,
+      "estimatedTotalHits": 4
+    }
+    "###);
+}
+
+#[actix_rt::test]
+async fn update_documents_with_geo_field() {
+    let server = Server::new().await;
+    let index = server.index("doggo");
+    index.update_settings(json!({"sortableAttributes": ["_geo"]})).await;
+
+    let documents = json!([
+        {
+            "id": "1",
+        },
+        {
+            "id": "2",
+            "_geo": null,
+        },
+        {
+            "id": "3",
+            "_geo": { "lat": 1, "lng": 1 },
+        },
+        {
+            "id": "4",
+            "_geo": { "lat": "1", "lng": "1" },
+        },
+    ]);
+
+    let (task, _status_code) = index.add_documents(documents, None).await;
+    let response = index.wait_task(task.uid()).await;
+    snapshot!(json_string!(response, { ".duration" => "[duration]", ".enqueuedAt" => "[date]", ".startedAt" => "[date]", ".finishedAt" => "[date]" }),
+        @r###"
+    {
+      "uid": 1,
+      "batchUid": 1,
+      "indexUid": "doggo",
+      "status": "succeeded",
+      "type": "documentAdditionOrUpdate",
+      "canceledBy": null,
+      "details": {
+        "receivedDocuments": 4,
+        "indexedDocuments": 4
+      },
+      "error": null,
+      "duration": "[duration]",
+      "enqueuedAt": "[date]",
+      "startedAt": "[date]",
+      "finishedAt": "[date]"
+    }
+    "###);
+
+    let (response, code) = index
+        .search_post(json!({"sort": ["_geoPoint(50.629973371633746,3.0569447399419567):desc"]}))
+        .await;
+    snapshot!(code, @"200 OK");
+    // we are expecting docs 4 and 3 first as they have geo
+    snapshot!(json_string!(response, { ".processingTimeMs" => "[time]" }),
+    @r###"
+    {
+      "hits": [
+        {
+          "id": "4",
+          "_geo": {
+            "lat": "1",
+            "lng": "1"
+          },
+          "_geoDistance": 5522018
+        },
+        {
+          "id": "3",
+          "_geo": {
+            "lat": 1,
+            "lng": 1
+          },
+          "_geoDistance": 5522018
+        },
+        {
+          "id": "1"
+        },
+        {
+          "id": "2",
+          "_geo": null
+        }
+      ],
+      "query": "",
+      "processingTimeMs": "[time]",
+      "limit": 20,
+      "offset": 0,
+      "estimatedTotalHits": 4
+    }
+    "###);
+
+    let updated_documents = json!([{
+      "id": "3",
+      "doggo": "kefir",
+    }]);
+    let (task, _status_code) = index.update_documents(updated_documents, None).await;
+    let response = index.wait_task(task.uid()).await;
+    snapshot!(json_string!(response, { ".duration" => "[duration]", ".enqueuedAt" => "[date]", ".startedAt" => "[date]", ".finishedAt" => "[date]" }),
+        @r###"
+    {
+      "uid": 2,
+      "batchUid": 2,
+      "indexUid": "doggo",
+      "status": "succeeded",
+      "type": "documentAdditionOrUpdate",
+      "canceledBy": null,
+      "details": {
+        "receivedDocuments": 1,
+        "indexedDocuments": 1
+      },
+      "error": null,
+      "duration": "[duration]",
+      "enqueuedAt": "[date]",
+      "startedAt": "[date]",
+      "finishedAt": "[date]"
+    }
+    "###);
+    let (response, code) = index.get_all_documents(GetAllDocumentsOptions::default()).await;
+
+    snapshot!(code, @"200 OK");
+    snapshot!(json_string!(response, { ".duration" => "[duration]", ".enqueuedAt" => "[date]", ".startedAt" => "[date]", ".finishedAt" => "[date]" }),
+    @r###"
+    {
+      "results": [
+        {
+          "id": "1"
+        },
+        {
+          "id": "2",
+          "_geo": null
+        },
+        {
+          "id": "3",
+          "_geo": {
+            "lat": 1,
+            "lng": 1
+          },
+          "doggo": "kefir"
+        },
+        {
+          "id": "4",
+          "_geo": {
+            "lat": "1",
+            "lng": "1"
+          }
+        }
+      ],
+      "offset": 0,
+      "limit": 20,
+      "total": 4
+    }
+    "###);
+
+    let (response, code) = index
+        .search_post(json!({"sort": ["_geoPoint(50.629973371633746,3.0569447399419567):desc"]}))
+        .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]" }),
+    @r###"
+    {
+      "hits": [
+        {
+          "id": "4",
+          "_geo": {
+            "lat": "1",
+            "lng": "1"
+          },
+          "_geoDistance": 5522018
+        },
+        {
+          "id": "3",
+          "_geo": {
+            "lat": 1,
+            "lng": 1
+          },
+          "doggo": "kefir",
+          "_geoDistance": 5522018
+        },
+        {
+          "id": "1"
+        },
+        {
+          "id": "2",
+          "_geo": null
+        }
+      ],
+      "query": "",
+      "processingTimeMs": "[time]",
+      "limit": 20,
+      "offset": 0,
+      "estimatedTotalHits": 4
+    }
+    "###);
 }
 
 #[actix_rt::test]

crates/meilisearch/tests/documents/delete_documents.rs

@@ -160,7 +160,11 @@ async fn delete_document_by_filter() {
     snapshot!(json_string!(stats), @r###"
     {
       "numberOfDocuments": 4,
+      "rawDocumentDbSize": 42,
+      "avgDocumentSize": 10,
       "isIndexing": false,
+      "numberOfEmbeddings": 0,
+      "numberOfEmbeddedDocuments": 0,
       "fieldDistribution": {
         "color": 3,
         "id": 4
@@ -207,7 +211,11 @@ async fn delete_document_by_filter() {
     snapshot!(json_string!(stats), @r###"
     {
       "numberOfDocuments": 2,
+      "rawDocumentDbSize": 16,
+      "avgDocumentSize": 8,
       "isIndexing": false,
+      "numberOfEmbeddings": 0,
+      "numberOfEmbeddedDocuments": 0,
       "fieldDistribution": {
         "color": 1,
         "id": 2
@@ -273,7 +281,11 @@ async fn delete_document_by_filter() {
     snapshot!(json_string!(stats), @r###"
     {
       "numberOfDocuments": 1,
+      "rawDocumentDbSize": 12,
+      "avgDocumentSize": 12,
       "isIndexing": false,
+      "numberOfEmbeddings": 0,
+      "numberOfEmbeddedDocuments": 0,
       "fieldDistribution": {
         "color": 1,
         "id": 1

crates/meilisearch/tests/dumps/mod.rs

@@ -27,9 +27,26 @@ async fn import_dump_v1_movie_raw() {
 
     let (stats, code) = index.stats().await;
     snapshot!(code, @"200 OK");
-    assert_eq!(
-        stats,
-        json!({ "numberOfDocuments": 53, "isIndexing": false, "fieldDistribution": {"genres": 53, "id": 53, "overview": 53, "poster": 53, "release_date": 53, "title": 53 }})
+    snapshot!(
+      json_string!(stats),
+      @r###"
+    {
+      "numberOfDocuments": 53,
+      "rawDocumentDbSize": 21965,
+      "avgDocumentSize": 414,
+      "isIndexing": false,
+      "numberOfEmbeddings": 0,
+      "numberOfEmbeddedDocuments": 0,
+      "fieldDistribution": {
+        "genres": 53,
+        "id": 53,
+        "overview": 53,
+        "poster": 53,
+        "release_date": 53,
+        "title": 53
+      }
+    }
+    "###
     );
 
     let (settings, code) = index.settings().await;
@@ -172,7 +189,11 @@ async fn import_dump_v1_movie_with_settings() {
         @r###"
     {
       "numberOfDocuments": 53,
+      "rawDocumentDbSize": 21965,
+      "avgDocumentSize": 414,
       "isIndexing": false,
+      "numberOfEmbeddings": 0,
+      "numberOfEmbeddedDocuments": 0,
       "fieldDistribution": {
         "genres": 53,
         "id": 53,
@@ -333,9 +354,26 @@ async fn import_dump_v1_rubygems_with_settings() {
 
     let (stats, code) = index.stats().await;
     snapshot!(code, @"200 OK");
-    assert_eq!(
-        stats,
-        json!({ "numberOfDocuments": 53, "isIndexing": false, "fieldDistribution": {"description": 53, "id": 53, "name": 53, "summary": 53, "total_downloads": 53, "version": 53 }})
+    snapshot!(
+      json_string!(stats),
+      @r###"
+    {
+      "numberOfDocuments": 53,
+      "rawDocumentDbSize": 8606,
+      "avgDocumentSize": 162,
+      "isIndexing": false,
+      "numberOfEmbeddings": 0,
+      "numberOfEmbeddedDocuments": 0,
+      "fieldDistribution": {
+        "description": 53,
+        "id": 53,
+        "name": 53,
+        "summary": 53,
+        "total_downloads": 53,
+        "version": 53
+      }
+    }
+    "###
     );
 
     let (settings, code) = index.settings().await;
@@ -483,9 +521,26 @@ async fn import_dump_v2_movie_raw() {
 
     let (stats, code) = index.stats().await;
     snapshot!(code, @"200 OK");
-    assert_eq!(
-        stats,
-        json!({ "numberOfDocuments": 53, "isIndexing": false, "fieldDistribution": {"genres": 53, "id": 53, "overview": 53, "poster": 53, "release_date": 53, "title": 53 }})
+    snapshot!(
+      json_string!(stats),
+      @r###"
+    {
+      "numberOfDocuments": 53,
+      "rawDocumentDbSize": 21965,
+      "avgDocumentSize": 414,
+      "isIndexing": false,
+      "numberOfEmbeddings": 0,
+      "numberOfEmbeddedDocuments": 0,
+      "fieldDistribution": {
+        "genres": 53,
+        "id": 53,
+        "overview": 53,
+        "poster": 53,
+        "release_date": 53,
+        "title": 53
+      }
+    }
+    "###
     );
 
     let (settings, code) = index.settings().await;
@@ -623,9 +678,26 @@ async fn import_dump_v2_movie_with_settings() {
 
     let (stats, code) = index.stats().await;
     snapshot!(code, @"200 OK");
-    assert_eq!(
-        stats,
-        json!({ "numberOfDocuments": 53, "isIndexing": false, "fieldDistribution": {"genres": 53, "id": 53, "overview": 53, "poster": 53, "release_date": 53, "title": 53 }})
+    snapshot!(
+      json_string!(stats),
+      @r###"
+    {
+      "numberOfDocuments": 53,
+      "rawDocumentDbSize": 21965,
+      "avgDocumentSize": 414,
+      "isIndexing": false,
+      "numberOfEmbeddings": 0,
+      "numberOfEmbeddedDocuments": 0,
+      "fieldDistribution": {
+        "genres": 53,
+        "id": 53,
+        "overview": 53,
+        "poster": 53,
+        "release_date": 53,
+        "title": 53
+      }
+    }
+    "###
     );
 
     let (settings, code) = index.settings().await;
@@ -773,9 +845,26 @@ async fn import_dump_v2_rubygems_with_settings() {
 
     let (stats, code) = index.stats().await;
     snapshot!(code, @"200 OK");
-    assert_eq!(
-        stats,
-        json!({ "numberOfDocuments": 53, "isIndexing": false, "fieldDistribution": {"description": 53, "id": 53, "name": 53, "summary": 53, "total_downloads": 53, "version": 53 }})
+    snapshot!(
+      json_string!(stats),
+      @r###"
+    {
+      "numberOfDocuments": 53,
+      "rawDocumentDbSize": 8606,
+      "avgDocumentSize": 162,
+      "isIndexing": false,
+      "numberOfEmbeddings": 0,
+      "numberOfEmbeddedDocuments": 0,
+      "fieldDistribution": {
+        "description": 53,
+        "id": 53,
+        "name": 53,
+        "summary": 53,
+        "total_downloads": 53,
+        "version": 53
+      }
+    }
+    "###
     );
 
     let (settings, code) = index.settings().await;
@@ -920,9 +1009,26 @@ async fn import_dump_v3_movie_raw() {
 
     let (stats, code) = index.stats().await;
     snapshot!(code, @"200 OK");
-    assert_eq!(
-        stats,
-        json!({ "numberOfDocuments": 53, "isIndexing": false, "fieldDistribution": {"genres": 53, "id": 53, "overview": 53, "poster": 53, "release_date": 53, "title": 53 }})
+    snapshot!(
+      json_string!(stats),
+      @r###"
+    {
+      "numberOfDocuments": 53,
+      "rawDocumentDbSize": 21965,
+      "avgDocumentSize": 414,
+      "isIndexing": false,
+      "numberOfEmbeddings": 0,
+      "numberOfEmbeddedDocuments": 0,
+      "fieldDistribution": {
+        "genres": 53,
+        "id": 53,
+        "overview": 53,
+        "poster": 53,
+        "release_date": 53,
+        "title": 53
+      }
+    }
+    "###
     );
 
     let (settings, code) = index.settings().await;
@@ -1060,9 +1166,26 @@ async fn import_dump_v3_movie_with_settings() {
 
     let (stats, code) = index.stats().await;
     snapshot!(code, @"200 OK");
-    assert_eq!(
-        stats,
-        json!({ "numberOfDocuments": 53, "isIndexing": false, "fieldDistribution": {"genres": 53, "id": 53, "overview": 53, "poster": 53, "release_date": 53, "title": 53 }})
+    snapshot!(
+      json_string!(stats),
+      @r###"
+    {
+      "numberOfDocuments": 53,
+      "rawDocumentDbSize": 21965,
+      "avgDocumentSize": 414,
+      "isIndexing": false,
+      "numberOfEmbeddings": 0,
+      "numberOfEmbeddedDocuments": 0,
+      "fieldDistribution": {
+        "genres": 53,
+        "id": 53,
+        "overview": 53,
+        "poster": 53,
+        "release_date": 53,
+        "title": 53
+      }
+    }
+    "###
     );
 
     let (settings, code) = index.settings().await;
@@ -1210,9 +1333,26 @@ async fn import_dump_v3_rubygems_with_settings() {
 
     let (stats, code) = index.stats().await;
     snapshot!(code, @"200 OK");
-    assert_eq!(
-        stats,
-        json!({ "numberOfDocuments": 53, "isIndexing": false, "fieldDistribution": {"description": 53, "id": 53, "name": 53, "summary": 53, "total_downloads": 53, "version": 53 }})
+    snapshot!(
+      json_string!(stats),
+      @r###"
+    {
+      "numberOfDocuments": 53,
+      "rawDocumentDbSize": 8606,
+      "avgDocumentSize": 162,
+      "isIndexing": false,
+      "numberOfEmbeddings": 0,
+      "numberOfEmbeddedDocuments": 0,
+      "fieldDistribution": {
+        "description": 53,
+        "id": 53,
+        "name": 53,
+        "summary": 53,
+        "total_downloads": 53,
+        "version": 53
+      }
+    }
+    "###
     );
 
     let (settings, code) = index.settings().await;
@@ -1357,9 +1497,26 @@ async fn import_dump_v4_movie_raw() {
 
     let (stats, code) = index.stats().await;
     snapshot!(code, @"200 OK");
-    assert_eq!(
-        stats,
-        json!({ "numberOfDocuments": 53, "isIndexing": false, "fieldDistribution": {"genres": 53, "id": 53, "overview": 53, "poster": 53, "release_date": 53, "title": 53 }})
+    snapshot!(
+      json_string!(stats),
+      @r###"
+    {
+      "numberOfDocuments": 53,
+      "rawDocumentDbSize": 21965,
+      "avgDocumentSize": 414,
+      "isIndexing": false,
+      "numberOfEmbeddings": 0,
+      "numberOfEmbeddedDocuments": 0,
+      "fieldDistribution": {
+        "genres": 53,
+        "id": 53,
+        "overview": 53,
+        "poster": 53,
+        "release_date": 53,
+        "title": 53
+      }
+    }
+    "###
     );
 
     let (settings, code) = index.settings().await;
@@ -1497,9 +1654,26 @@ async fn import_dump_v4_movie_with_settings() {
 
     let (stats, code) = index.stats().await;
     snapshot!(code, @"200 OK");
-    assert_eq!(
-        stats,
-        json!({ "numberOfDocuments": 53, "isIndexing": false, "fieldDistribution": {"genres": 53, "id": 53, "overview": 53, "poster": 53, "release_date": 53, "title": 53 }})
+    snapshot!(
+      json_string!(stats),
+      @r###"
+    {
+      "numberOfDocuments": 53,
+      "rawDocumentDbSize": 21965,
+      "avgDocumentSize": 414,
+      "isIndexing": false,
+      "numberOfEmbeddings": 0,
+      "numberOfEmbeddedDocuments": 0,
+      "fieldDistribution": {
+        "genres": 53,
+        "id": 53,
+        "overview": 53,
+        "poster": 53,
+        "release_date": 53,
+        "title": 53
+      }
+    }
+    "###
     );
 
     let (settings, code) = index.settings().await;
@@ -1647,9 +1821,26 @@ async fn import_dump_v4_rubygems_with_settings() {
 
     let (stats, code) = index.stats().await;
     snapshot!(code, @"200 OK");
-    assert_eq!(
-        stats,
-        json!({ "numberOfDocuments": 53, "isIndexing": false, "fieldDistribution": {"description": 53, "id": 53, "name": 53, "summary": 53, "total_downloads": 53, "version": 53 }})
+    snapshot!(
+      json_string!(stats),
+      @r###"
+    {
+      "numberOfDocuments": 53,
+      "rawDocumentDbSize": 8606,
+      "avgDocumentSize": 162,
+      "isIndexing": false,
+      "numberOfEmbeddings": 0,
+      "numberOfEmbeddedDocuments": 0,
+      "fieldDistribution": {
+        "description": 53,
+        "id": 53,
+        "name": 53,
+        "summary": 53,
+        "total_downloads": 53,
+        "version": 53
+      }
+    }
+    "###
     );
 
     let (settings, code) = index.settings().await;
@@ -1798,33 +1989,37 @@ async fn import_dump_v5() {
         server.wait_task(task["uid"].as_u64().unwrap()).await;
     }
 
-    let expected_stats = json!({
-        "numberOfDocuments": 10,
-        "isIndexing": false,
-        "fieldDistribution": {
-            "cast": 10,
-            "director": 10,
-            "genres": 10,
-            "id": 10,
-            "overview": 10,
-            "popularity": 10,
-            "poster_path": 10,
-            "producer": 10,
-            "production_companies": 10,
-            "release_date": 10,
-            "tagline": 10,
-            "title": 10,
-            "vote_average": 10,
-            "vote_count": 10
-        }
-    });
-
     let index1 = server.index("test");
     let index2 = server.index("test2");
 
     let (stats, code) = index1.stats().await;
     snapshot!(code, @"200 OK");
-    assert_eq!(stats, expected_stats);
+    snapshot!(json_string!(stats), @r###"
+    {
+      "numberOfDocuments": 10,
+      "rawDocumentDbSize": 6782,
+      "avgDocumentSize": 678,
+      "isIndexing": false,
+      "numberOfEmbeddings": 0,
+      "numberOfEmbeddedDocuments": 0,
+      "fieldDistribution": {
+        "cast": 10,
+        "director": 10,
+        "genres": 10,
+        "id": 10,
+        "overview": 10,
+        "popularity": 10,
+        "poster_path": 10,
+        "producer": 10,
+        "production_companies": 10,
+        "release_date": 10,
+        "tagline": 10,
+        "title": 10,
+        "vote_average": 10,
+        "vote_count": 10
+      }
+    }
+    "###);
 
     let (docs, code) = index2.get_all_documents(GetAllDocumentsOptions::default()).await;
     snapshot!(code, @"200 OK");
@@ -1835,7 +2030,34 @@ async fn import_dump_v5() {
 
     let (stats, code) = index2.stats().await;
     snapshot!(code, @"200 OK");
-    assert_eq!(stats, expected_stats);
+    snapshot!(
+      json_string!(stats),
+      @r###"
+    {
+      "numberOfDocuments": 10,
+      "rawDocumentDbSize": 6782,
+      "avgDocumentSize": 678,
+      "isIndexing": false,
+      "numberOfEmbeddings": 0,
+      "numberOfEmbeddedDocuments": 0,
+      "fieldDistribution": {
+        "cast": 10,
+        "director": 10,
+        "genres": 10,
+        "id": 10,
+        "overview": 10,
+        "popularity": 10,
+        "poster_path": 10,
+        "producer": 10,
+        "production_companies": 10,
+        "release_date": 10,
+        "tagline": 10,
+        "title": 10,
+        "vote_average": 10,
+        "vote_count": 10
+      }
+    }
+    "###);
 
     let (keys, code) = server.list_api_keys("").await;
     snapshot!(code, @"200 OK");

crates/meilisearch/tests/search/mod.rs

@@ -128,6 +128,40 @@ async fn search_with_stop_word() {
           .await;
 }
 
+#[actix_rt::test]
+async fn search_with_typo_settings() {
+    // related to https://github.com/meilisearch/meilisearch/issues/5240
+    let server = Server::new().await;
+    let index = server.index("test");
+
+    let (_, code) = index
+        .update_settings(json!({"typoTolerance": { "disableOnAttributes": ["title", "id"]}}))
+        .await;
+    meili_snap::snapshot!(code, @"202 Accepted");
+
+    let documents = DOCUMENTS.clone();
+    let (task, _status_code) = index.add_documents(documents, None).await;
+    index.wait_task(task.uid()).await.succeeded();
+
+    index
+        .search(json!({"q": "287947" }), |response, code| {
+            assert_eq!(code, 200, "{}", response);
+            snapshot!(json_string!(response["hits"]), @r###"
+            [
+              {
+                "title": "Shazam!",
+                "id": "287947",
+                "color": [
+                  "green",
+                  "blue"
+                ]
+              }
+            ]
+            "###);
+        })
+        .await;
+}
+
 #[actix_rt::test]
 async fn phrase_search_with_stop_word() {
     // related to https://github.com/meilisearch/meilisearch/issues/3521

crates/meilisearch/tests/stats/mod.rs

@@ -1,3 +1,4 @@
+use meili_snap::{json_string, snapshot};
 use time::format_description::well_known::Rfc3339;
 use time::OffsetDateTime;
 
@@ -74,3 +75,269 @@ async fn stats() {
     assert_eq!(response["indexes"]["test"]["fieldDistribution"]["name"], 1);
     assert_eq!(response["indexes"]["test"]["fieldDistribution"]["age"], 1);
 }
+
+#[actix_rt::test]
+async fn add_remove_embeddings() {
+    let server = Server::new().await;
+    let index = server.index("doggo");
+
+    let (response, code) = index
+        .update_settings(json!({
+          "embedders": {
+            "manual": {
+                "source": "userProvided",
+                "dimensions": 3,
+            },
+            "handcrafted": {
+                "source": "userProvided",
+                "dimensions": 3,
+            },
+
+          },
+        }))
+        .await;
+    snapshot!(code, @"202 Accepted");
+    server.wait_task(response.uid()).await.succeeded();
+
+    // 2 embedded documents for 5 embeddings in total
+    let documents = json!([
+      {"id": 0, "name": "kefir", "_vectors": { "manual": [0, 0, 0], "handcrafted": [0, 0, 0] }},
+      {"id": 1, "name": "echo", "_vectors": { "manual": [1, 1, 1], "handcrafted": [[1, 1, 1], [2, 2, 2]] }},
+    ]);
+
+    let (response, code) = index.add_documents(documents, None).await;
+    snapshot!(code, @"202 Accepted");
+    index.wait_task(response.uid()).await.succeeded();
+
+    let (stats, _code) = index.stats().await;
+    snapshot!(json_string!(stats), @r###"
+    {
+      "numberOfDocuments": 2,
+      "rawDocumentDbSize": 27,
+      "avgDocumentSize": 13,
+      "isIndexing": false,
+      "numberOfEmbeddings": 5,
+      "numberOfEmbeddedDocuments": 2,
+      "fieldDistribution": {
+        "id": 2,
+        "name": 2
+      }
+    }
+    "###);
+
+    // 2 embedded documents for 3 embeddings in total
+    let documents = json!([
+      {"id": 1, "name": "echo", "_vectors": { "manual": [1, 1, 1], "handcrafted": null }},
+    ]);
+
+    let (response, code) = index.update_documents(documents, None).await;
+    snapshot!(code, @"202 Accepted");
+    index.wait_task(response.uid()).await.succeeded();
+
+    let (stats, _code) = index.stats().await;
+    snapshot!(json_string!(stats), @r###"
+    {
+      "numberOfDocuments": 2,
+      "rawDocumentDbSize": 27,
+      "avgDocumentSize": 13,
+      "isIndexing": false,
+      "numberOfEmbeddings": 3,
+      "numberOfEmbeddedDocuments": 2,
+      "fieldDistribution": {
+        "id": 2,
+        "name": 2
+      }
+    }
+    "###);
+
+    // 2 embedded documents for 2 embeddings in total
+    let documents = json!([
+        {"id": 0, "name": "kefir", "_vectors": { "manual": null, "handcrafted": [0, 0, 0] }},
+    ]);
+
+    let (response, code) = index.update_documents(documents, None).await;
+    snapshot!(code, @"202 Accepted");
+    index.wait_task(response.uid()).await.succeeded();
+
+    let (stats, _code) = index.stats().await;
+    snapshot!(json_string!(stats), @r###"
+    {
+      "numberOfDocuments": 2,
+      "rawDocumentDbSize": 27,
+      "avgDocumentSize": 13,
+      "isIndexing": false,
+      "numberOfEmbeddings": 2,
+      "numberOfEmbeddedDocuments": 2,
+      "fieldDistribution": {
+        "id": 2,
+        "name": 2
+      }
+    }
+    "###);
+
+    // 1 embedded documents for 2 embeddings in total
+    let documents = json!([
+        {"id": 0, "name": "kefir", "_vectors": { "manual": [0, 0, 0], "handcrafted": [0, 0, 0] }},
+        {"id": 1, "name": "echo", "_vectors": { "manual": null, "handcrafted": null }},
+    ]);
+
+    let (response, code) = index.update_documents(documents, None).await;
+    snapshot!(code, @"202 Accepted");
+    index.wait_task(response.uid()).await.succeeded();
+
+    let (stats, _code) = index.stats().await;
+    snapshot!(json_string!(stats), @r###"
+    {
+      "numberOfDocuments": 2,
+      "rawDocumentDbSize": 27,
+      "avgDocumentSize": 13,
+      "isIndexing": false,
+      "numberOfEmbeddings": 2,
+      "numberOfEmbeddedDocuments": 1,
+      "fieldDistribution": {
+        "id": 2,
+        "name": 2
+      }
+    }
+    "###);
+}
+
+#[actix_rt::test]
+async fn add_remove_embedded_documents() {
+    let server = Server::new().await;
+    let index = server.index("doggo");
+
+    let (response, code) = index
+        .update_settings(json!({
+          "embedders": {
+            "manual": {
+                "source": "userProvided",
+                "dimensions": 3,
+            },
+            "handcrafted": {
+                "source": "userProvided",
+                "dimensions": 3,
+            },
+
+          },
+        }))
+        .await;
+    snapshot!(code, @"202 Accepted");
+    server.wait_task(response.uid()).await.succeeded();
+
+    // 2 embedded documents for 5 embeddings in total
+    let documents = json!([
+      {"id": 0, "name": "kefir", "_vectors": { "manual": [0, 0, 0], "handcrafted": [0, 0, 0] }},
+      {"id": 1, "name": "echo", "_vectors": { "manual": [1, 1, 1], "handcrafted": [[1, 1, 1], [2, 2, 2]] }},
+    ]);
+
+    let (response, code) = index.add_documents(documents, None).await;
+    snapshot!(code, @"202 Accepted");
+    index.wait_task(response.uid()).await.succeeded();
+
+    let (stats, _code) = index.stats().await;
+    snapshot!(json_string!(stats), @r###"
+    {
+      "numberOfDocuments": 2,
+      "rawDocumentDbSize": 27,
+      "avgDocumentSize": 13,
+      "isIndexing": false,
+      "numberOfEmbeddings": 5,
+      "numberOfEmbeddedDocuments": 2,
+      "fieldDistribution": {
+        "id": 2,
+        "name": 2
+      }
+    }
+    "###);
+
+    // delete one embedded document, remaining 1 embedded documents for 3 embeddings in total
+    let (response, code) = index.delete_document(0).await;
+    snapshot!(code, @"202 Accepted");
+    index.wait_task(response.uid()).await.succeeded();
+
+    let (stats, _code) = index.stats().await;
+    snapshot!(json_string!(stats), @r###"
+    {
+      "numberOfDocuments": 1,
+      "rawDocumentDbSize": 13,
+      "avgDocumentSize": 13,
+      "isIndexing": false,
+      "numberOfEmbeddings": 3,
+      "numberOfEmbeddedDocuments": 1,
+      "fieldDistribution": {
+        "id": 1,
+        "name": 1
+      }
+    }
+    "###);
+}
+
+#[actix_rt::test]
+async fn update_embedder_settings() {
+    let server = Server::new().await;
+    let index = server.index("doggo");
+
+    // 2 embedded documents for 3 embeddings in total
+    // but no embedders are added in the settings yet so we expect 0 embedded documents for 0 embeddings in total
+    let documents = json!([
+      {"id": 0, "name": "kefir", "_vectors": { "manual": [0, 0, 0], "handcrafted": [0, 0, 0] }},
+      {"id": 1, "name": "echo", "_vectors": { "manual": [1, 1, 1], "handcrafted": null }},
+    ]);
+
+    let (response, code) = index.add_documents(documents, None).await;
+    snapshot!(code, @"202 Accepted");
+    index.wait_task(response.uid()).await.succeeded();
+
+    let (stats, _code) = index.stats().await;
+    snapshot!(json_string!(stats), @r###"
+    {
+      "numberOfDocuments": 2,
+      "rawDocumentDbSize": 108,
+      "avgDocumentSize": 54,
+      "isIndexing": false,
+      "numberOfEmbeddings": 0,
+      "numberOfEmbeddedDocuments": 0,
+      "fieldDistribution": {
+        "id": 2,
+        "name": 2
+      }
+    }
+    "###);
+
+    // add embedders to the settings
+    // 2 embedded documents for 3 embeddings in total
+    let (response, code) = index
+        .update_settings(json!({
+          "embedders": {
+            "manual": {
+                "source": "userProvided",
+                "dimensions": 3,
+            },
+            "handcrafted": {
+                "source": "userProvided",
+                "dimensions": 3,
+            },
+
+          },
+        }))
+        .await;
+    snapshot!(code, @"202 Accepted");
+    server.wait_task(response.uid()).await.succeeded();
+
+    let (stats, _code) = index.stats().await;
+    snapshot!(json_string!(stats), @r###"
+    {
+      "numberOfDocuments": 2,
+      "rawDocumentDbSize": 108,
+      "avgDocumentSize": 54,
+      "isIndexing": false,
+      "numberOfEmbeddings": 3,
+      "numberOfEmbeddedDocuments": 2,
+      "fieldDistribution": {
+        "id": 2,
+        "name": 2
+      }
+    }
+    "###);
+}

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 v1.13.0");
+    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 v1.13.2");
 }
 
 #[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 1.13.1 is higher than the Meilisearch version 1.13.0. Downgrade is not supported");
+    snapshot!(err, @"Database version 1.13.3 is higher than the Meilisearch version 1.13.2. Downgrade is not supported");
 }
 
 #[actix_rt::test]

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": "v1.13.0"
+        "upgradeTo": "v1.13.2"
       },
       "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": "v1.13.0"
+        "upgradeTo": "v1.13.2"
       },
       "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": "v1.13.0"
+        "upgradeTo": "v1.13.2"
       },
       "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": "v1.13.0"
+        "upgradeTo": "v1.13.2"
       },
       "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": "v1.13.0"
+        "upgradeTo": "v1.13.2"
       },
       "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": "v1.13.0"
+        "upgradeTo": "v1.13.2"
       },
       "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

@@ -1,6 +1,5 @@
 ---
 source: crates/meilisearch/tests/upgrade/v1_12/v1_12_0.rs
-snapshot_kind: text
 ---
 {
   "results": [
@@ -9,7 +8,7 @@ snapshot_kind: text
       "progress": null,
       "details": {
         "upgradeFrom": "v1.12.0",
-        "upgradeTo": "v1.13.0"
+        "upgradeTo": "v1.13.2"
       },
       "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

@@ -1,6 +1,5 @@
 ---
 source: crates/meilisearch/tests/upgrade/v1_12/v1_12_0.rs
-snapshot_kind: text
 ---
 {
   "results": [
@@ -13,7 +12,7 @@ snapshot_kind: text
       "canceledBy": null,
       "details": {
         "upgradeFrom": "v1.12.0",
-        "upgradeTo": "v1.13.0"
+        "upgradeTo": "v1.13.2"
       },
       "error": null,
       "duration": "[duration]",

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

@@ -134,7 +134,11 @@ async fn check_the_index_scheduler(server: &Server) {
       "indexes": {
         "kefir": {
           "numberOfDocuments": 1,
+          "rawDocumentDbSize": 109,
+          "avgDocumentSize": 109,
           "isIndexing": false,
+          "numberOfEmbeddings": 0,
+          "numberOfEmbeddedDocuments": 0,
           "fieldDistribution": {
             "age": 1,
             "description": 1,
@@ -214,7 +218,11 @@ async fn check_the_index_scheduler(server: &Server) {
       "indexes": {
         "kefir": {
           "numberOfDocuments": 1,
+          "rawDocumentDbSize": 109,
+          "avgDocumentSize": 109,
           "isIndexing": false,
+          "numberOfEmbeddings": 0,
+          "numberOfEmbeddedDocuments": 0,
           "fieldDistribution": {
             "age": 1,
             "description": 1,
@@ -228,10 +236,14 @@ async fn check_the_index_scheduler(server: &Server) {
     "###);
     let index = server.index("kefir");
     let (stats, _) = index.stats().await;
-    snapshot!(stats, @r#"
+    snapshot!(stats, @r###"
     {
       "numberOfDocuments": 1,
+      "rawDocumentDbSize": 109,
+      "avgDocumentSize": 109,
       "isIndexing": false,
+      "numberOfEmbeddings": 0,
+      "numberOfEmbeddedDocuments": 0,
       "fieldDistribution": {
         "age": 1,
         "description": 1,
@@ -240,7 +252,7 @@ async fn check_the_index_scheduler(server: &Server) {
         "surname": 1
       }
     }
-    "#);
+    "###);
 
     // Delete all the tasks of a specific batch
     let (task, _) = server.delete_tasks("batchUids=10").await;

crates/milli/Cargo.toml

@@ -85,7 +85,7 @@ rhai = { git = "https://github.com/rhaiscript/rhai", rev = "ef3df63121d27aacd838
     "no_time",
     "sync",
 ] }
-arroy = "0.5.0"
+arroy = { git = "https://github.com/meilisearch/arroy", rev = "55fb0e8006f4f00ad3197b06bda348133f6ffffa" }
 rand = "0.8.5"
 tracing = "0.1.41"
 ureq = { version = "2.12.1", features = ["json"] }

crates/milli/src/database_stats.rs

@@ -0,0 +1,96 @@
+use heed::types::Bytes;
+use heed::Database;
+use heed::RoTxn;
+use serde::{Deserialize, Serialize};
+
+#[derive(Debug, Clone, Copy, Serialize, Deserialize, PartialEq, Eq, Default)]
+#[serde(rename_all = "camelCase")]
+/// The stats of a database.
+pub struct DatabaseStats {
+    /// The number of entries in the database.
+    number_of_entries: u64,
+    /// The total size of the keys in the database.
+    total_key_size: u64,
+    /// The total size of the values in the database.
+    total_value_size: u64,
+}
+
+impl DatabaseStats {
+    /// Returns the stats of the database.
+    ///
+    /// This function iterates over the whole database and computes the stats.
+    /// It is not efficient and should be cached somewhere.
+    pub(crate) fn new(database: Database<Bytes, Bytes>, rtxn: &RoTxn<'_>) -> heed::Result<Self> {
+        let mut database_stats =
+            Self { number_of_entries: 0, total_key_size: 0, total_value_size: 0 };
+
+        let mut iter = database.iter(rtxn)?;
+        while let Some((key, value)) = iter.next().transpose()? {
+            let key_size = key.len() as u64;
+            let value_size = value.len() as u64;
+            database_stats.total_key_size += key_size;
+            database_stats.total_value_size += value_size;
+        }
+
+        database_stats.number_of_entries = database.len(rtxn)?;
+
+        Ok(database_stats)
+    }
+
+    /// Recomputes the stats of the database and returns the new stats.
+    ///
+    /// This function is used to update the stats of the database when some keys are modified.
+    /// It is more efficient than the `new` function because it does not iterate over the whole database but only the modified keys comparing the before and after states.
+    pub(crate) fn recompute<I, K>(
+        mut stats: Self,
+        database: Database<Bytes, Bytes>,
+        before_rtxn: &RoTxn<'_>,
+        after_rtxn: &RoTxn<'_>,
+        modified_keys: I,
+    ) -> heed::Result<Self>
+    where
+        I: IntoIterator<Item = K>,
+        K: AsRef<[u8]>,
+    {
+        for key in modified_keys {
+            let key = key.as_ref();
+            if let Some(value) = database.get(after_rtxn, key)? {
+                let key_size = key.len() as u64;
+                let value_size = value.len() as u64;
+                stats.total_key_size = stats.total_key_size.saturating_add(key_size);
+                stats.total_value_size = stats.total_value_size.saturating_add(value_size);
+            }
+
+            if let Some(value) = database.get(before_rtxn, key)? {
+                let key_size = key.len() as u64;
+                let value_size = value.len() as u64;
+                stats.total_key_size = stats.total_key_size.saturating_sub(key_size);
+                stats.total_value_size = stats.total_value_size.saturating_sub(value_size);
+            }
+        }
+
+        stats.number_of_entries = database.len(after_rtxn)?;
+
+        Ok(stats)
+    }
+
+    pub fn average_key_size(&self) -> u64 {
+        self.total_key_size.checked_div(self.number_of_entries).unwrap_or(0)
+    }
+
+    pub fn average_value_size(&self) -> u64 {
+        self.total_value_size.checked_div(self.number_of_entries).unwrap_or(0)
+    }
+
+    pub fn number_of_entries(&self) -> u64 {
+        self.number_of_entries
+    }
+
+    pub fn total_key_size(&self) -> u64 {
+        self.total_key_size
+    }
+
+    pub fn total_value_size(&self) -> u64 {
+        self.total_value_size
+    }
+}

crates/milli/src/index.rs

@@ -11,6 +11,7 @@ use rstar::RTree;
 use serde::{Deserialize, Serialize};
 
 use crate::constants::{self, RESERVED_VECTORS_FIELD_NAME};
+use crate::database_stats::DatabaseStats;
 use crate::documents::PrimaryKey;
 use crate::error::{InternalError, UserError};
 use crate::fields_ids_map::FieldsIdsMap;
@@ -22,7 +23,7 @@ use crate::heed_codec::version::VersionCodec;
 use crate::heed_codec::{BEU16StrCodec, FstSetCodec, StrBEU16Codec, StrRefCodec};
 use crate::order_by_map::OrderByMap;
 use crate::proximity::ProximityPrecision;
-use crate::vector::{ArroyWrapper, Embedding, EmbeddingConfig};
+use crate::vector::{ArroyStats, ArroyWrapper, Embedding, EmbeddingConfig};
 use crate::{
     default_criteria, CboRoaringBitmapCodec, Criterion, DocumentId, ExternalDocumentsIds,
     FacetDistribution, FieldDistribution, FieldId, FieldIdMapMissingEntry, FieldIdWordCountCodec,
@@ -74,6 +75,7 @@ pub mod main_key {
     pub const LOCALIZED_ATTRIBUTES_RULES: &str = "localized_attributes_rules";
     pub const FACET_SEARCH: &str = "facet_search";
     pub const PREFIX_SEARCH: &str = "prefix_search";
+    pub const DOCUMENTS_STATS: &str = "documents_stats";
 }
 
 pub mod db_name {
@@ -403,6 +405,58 @@ impl Index {
         Ok(count.unwrap_or_default())
     }
 
+    /// Updates the stats of the documents database based on the previous stats and the modified docids.
+    pub fn update_documents_stats(
+        &self,
+        wtxn: &mut RwTxn<'_>,
+        modified_docids: roaring::RoaringBitmap,
+    ) -> Result<()> {
+        let before_rtxn = self.read_txn()?;
+        let document_stats = match self.documents_stats(&before_rtxn)? {
+            Some(before_stats) => DatabaseStats::recompute(
+                before_stats,
+                self.documents.remap_types(),
+                &before_rtxn,
+                wtxn,
+                modified_docids.iter().map(|docid| docid.to_be_bytes()),
+            )?,
+            None => {
+                // This should never happen when there are already documents in the index, the documents stats should be present.
+                // If it happens, it means that the index was not properly initialized/upgraded.
+                debug_assert_eq!(
+                    self.documents.len(&before_rtxn)?,
+                    0,
+                    "The documents stats should be present when there are documents in the index"
+                );
+                tracing::warn!("No documents stats found, creating new ones");
+                DatabaseStats::new(self.documents.remap_types(), &*wtxn)?
+            }
+        };
+
+        self.put_documents_stats(wtxn, document_stats)?;
+        Ok(())
+    }
+
+    /// Writes the stats of the documents database.
+    pub fn put_documents_stats(
+        &self,
+        wtxn: &mut RwTxn<'_>,
+        stats: DatabaseStats,
+    ) -> heed::Result<()> {
+        self.main.remap_types::<Str, SerdeJson<DatabaseStats>>().put(
+            wtxn,
+            main_key::DOCUMENTS_STATS,
+            &stats,
+        )
+    }
+
+    /// Returns the stats of the documents database.
+    pub fn documents_stats(&self, rtxn: &RoTxn<'_>) -> heed::Result<Option<DatabaseStats>> {
+        self.main
+            .remap_types::<Str, SerdeJson<DatabaseStats>>()
+            .get(rtxn, main_key::DOCUMENTS_STATS)
+    }
+
     /* primary key */
 
     /// Writes the documents primary key, this is the field name that is used to store the id.
@@ -1731,6 +1785,18 @@ impl Index {
         let compute_prefixes = self.prefix_search(rtxn)?.unwrap_or_default();
         Ok(PrefixSettings { compute_prefixes, max_prefix_length: 4, prefix_count_threshold: 100 })
     }
+
+    pub fn arroy_stats(&self, rtxn: &RoTxn<'_>) -> Result<ArroyStats> {
+        let mut stats = ArroyStats::default();
+        let embedding_configs = self.embedding_configs(rtxn)?;
+        for config in embedding_configs {
+            let embedder_id = self.embedder_category_id.get(rtxn, &config.name)?.unwrap();
+            let reader =
+                ArroyWrapper::new(self.vector_arroy, embedder_id, config.config.quantized());
+            reader.aggregate_stats(rtxn, &mut stats)?;
+        }
+        Ok(stats)
+    }
 }
 
 #[derive(Debug, Deserialize, Serialize)]

crates/milli/src/lib.rs

@@ -10,6 +10,7 @@ pub mod documents;
 
 mod asc_desc;
 mod criterion;
+pub mod database_stats;
 mod error;
 mod external_documents_ids;
 pub mod facet;

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

@@ -215,7 +215,7 @@ pub fn partially_initialized_term_from_word(
     let mut zero_typo = None;
     let mut prefix_of = BTreeSet::new();
 
-    if fst.contains(word) {
+    if fst.contains(word) || ctx.index.exact_word_docids.get(ctx.txn, word)?.is_some() {
         zero_typo = Some(word_interned);
     }
 

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

@@ -307,6 +307,7 @@ where
         let current_span = tracing::Span::current();
 
         // Run extraction pipeline in parallel.
+        let mut modified_docids = RoaringBitmap::new();
         pool.install(|| {
                 let settings_diff_cloned = settings_diff.clone();
                 rayon::spawn(move || {
@@ -367,7 +368,7 @@ where
                         Err(status) => {
                             if let Some(typed_chunks) = chunk_accumulator.pop_longest() {
                                 let (docids, is_merged_database) =
-                                    write_typed_chunk_into_index(self.wtxn, self.index, &settings_diff, typed_chunks)?;
+                                    write_typed_chunk_into_index(self.wtxn, self.index, &settings_diff, typed_chunks, &mut modified_docids)?;
                                 if !docids.is_empty() {
                                     final_documents_ids |= docids;
                                     let documents_seen_count = final_documents_ids.len();
@@ -467,6 +468,10 @@ where
                 Ok(())
             }).map_err(InternalError::from)??;
 
+        if !settings_diff.settings_update_only {
+            // Update the stats of the documents database when there is a document update.
+            self.index.update_documents_stats(self.wtxn, modified_docids)?;
+        }
         // We write the field distribution into the main database
         self.index.put_field_distribution(self.wtxn, &field_distribution)?;
 

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

@@ -129,6 +129,7 @@ pub(crate) fn write_typed_chunk_into_index(
     index: &Index,
     settings_diff: &InnerIndexSettingsDiff,
     typed_chunks: Vec<TypedChunk>,
+    modified_docids: &mut RoaringBitmap,
 ) -> Result<(RoaringBitmap, bool)> {
     let mut is_merged_database = false;
     match typed_chunks[0] {
@@ -214,6 +215,7 @@ pub(crate) fn write_typed_chunk_into_index(
                         kind: DocumentOperationKind::Create,
                     });
                     docids.insert(docid);
+                    modified_docids.insert(docid);
                 } else {
                     db.delete(wtxn, &docid)?;
                     operations.push(DocumentOperation {
@@ -222,6 +224,7 @@ pub(crate) fn write_typed_chunk_into_index(
                         kind: DocumentOperationKind::Delete,
                     });
                     docids.remove(docid);
+                    modified_docids.insert(docid);
                 }
             }
             let external_documents_docids = index.external_documents_ids();

crates/milli/src/update/new/document_change.rs

@@ -144,7 +144,7 @@ impl<'doc> Update<'doc> {
         )?)
     }
 
-    pub fn updated(&self) -> DocumentFromVersions<'_, 'doc> {
+    pub fn only_changed_fields(&self) -> DocumentFromVersions<'_, 'doc> {
         DocumentFromVersions::new(&self.new)
     }
 
@@ -182,7 +182,7 @@ impl<'doc> Update<'doc> {
         let mut cached_current = None;
         let mut updated_selected_field_count = 0;
 
-        for entry in self.updated().iter_top_level_fields() {
+        for entry in self.only_changed_fields().iter_top_level_fields() {
             let (key, updated_value) = entry?;
 
             if perm_json_p::select_field(key, fields, &[]) == perm_json_p::Selection::Skip {
@@ -241,7 +241,7 @@ impl<'doc> Update<'doc> {
         Ok(has_deleted_fields)
     }
 
-    pub fn updated_vectors(
+    pub fn only_changed_vectors(
         &self,
         doc_alloc: &'doc Bump,
         embedders: &'doc EmbeddingConfigs,

crates/milli/src/update/new/extract/cache.rs

@@ -711,15 +711,17 @@ impl DelAddRoaringBitmap {
         DelAddRoaringBitmap { del, add }
     }
 
-    pub fn apply_to(&self, documents_ids: &mut RoaringBitmap) {
+    pub fn apply_to(&self, documents_ids: &mut RoaringBitmap, modified_docids: &mut RoaringBitmap) {
         let DelAddRoaringBitmap { del, add } = self;
 
         if let Some(del) = del {
             *documents_ids -= del;
+            *modified_docids |= del;
         }
 
         if let Some(add) = add {
             *documents_ids |= add;
+            *modified_docids |= add;
         }
     }
 }

crates/milli/src/update/new/extract/geo/mod.rs

@@ -199,7 +199,7 @@ impl<'extractor> Extractor<'extractor> for GeoExtractor {
                         .transpose()?;
 
                     let updated_geo = update
-                        .updated()
+                        .merged(rtxn, index, db_fields_ids_map)?
                         .geo_field()?
                         .map(|geo| extract_geo_coordinates(external_id, geo))
                         .transpose()?;

crates/milli/src/update/new/extract/vectors/mod.rs

@@ -99,7 +99,8 @@ impl<'a, 'b, 'extractor> Extractor<'extractor> for EmbeddingExtractor<'a, 'b> {
                         context.db_fields_ids_map,
                         &context.doc_alloc,
                     )?;
-                    let new_vectors = update.updated_vectors(&context.doc_alloc, self.embedders)?;
+                    let new_vectors =
+                        update.only_changed_vectors(&context.doc_alloc, self.embedders)?;
 
                     if let Some(new_vectors) = &new_vectors {
                         unused_vectors_distribution.append(new_vectors)?;

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

@@ -32,6 +32,7 @@ pub(super) fn extract_all<'pl, 'extractor, DC, MSP>(
     field_distribution: &mut BTreeMap<String, u64>,
     mut index_embeddings: Vec<IndexEmbeddingConfig>,
     document_ids: &mut RoaringBitmap,
+    modified_docids: &mut RoaringBitmap,
 ) -> Result<(FacetFieldIdsDelta, Vec<IndexEmbeddingConfig>)>
 where
     DC: DocumentChanges<'pl>,
@@ -70,7 +71,7 @@ where
                 // adding the delta should never cause a negative result, as we are removing fields that previously existed.
                 *current = current.saturating_add_signed(delta);
             }
-            document_extractor_data.docids_delta.apply_to(document_ids);
+            document_extractor_data.docids_delta.apply_to(document_ids, modified_docids);
         }
 
         field_distribution.retain(|_, v| *v != 0);
@@ -256,7 +257,7 @@ where
                     let Some(deladd) = data.remove(&config.name) else {
                         continue 'data;
                     };
-                    deladd.apply_to(&mut config.user_provided);
+                    deladd.apply_to(&mut config.user_provided, modified_docids);
                 }
             }
         }

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

@@ -129,6 +129,7 @@ where
     let index_embeddings = index.embedding_configs(wtxn)?;
     let mut field_distribution = index.field_distribution(wtxn)?;
     let mut document_ids = index.documents_ids(wtxn)?;
+    let mut modified_docids = roaring::RoaringBitmap::new();
 
     thread::scope(|s| -> Result<()> {
         let indexer_span = tracing::Span::current();
@@ -137,6 +138,7 @@ where
         // prevent moving the field_distribution and document_ids in the inner closure...
         let field_distribution = &mut field_distribution;
         let document_ids = &mut document_ids;
+        let modified_docids = &mut modified_docids;
         let extractor_handle =
             Builder::new().name(S("indexer-extractors")).spawn_scoped(s, move || {
                 pool.install(move || {
@@ -151,6 +153,7 @@ where
                         field_distribution,
                         index_embeddings,
                         document_ids,
+                        modified_docids,
                     )
                 })
                 .unwrap()
@@ -225,6 +228,7 @@ where
         embedders,
         field_distribution,
         document_ids,
+        modified_docids,
     )?;
 
     Ok(())

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

@@ -113,6 +113,7 @@ where
     Ok(())
 }
 
+#[allow(clippy::too_many_arguments)]
 pub(super) fn update_index(
     index: &Index,
     wtxn: &mut RwTxn<'_>,
@@ -121,6 +122,7 @@ pub(super) fn update_index(
     embedders: EmbeddingConfigs,
     field_distribution: std::collections::BTreeMap<String, u64>,
     document_ids: roaring::RoaringBitmap,
+    modified_docids: roaring::RoaringBitmap,
 ) -> Result<()> {
     index.put_fields_ids_map(wtxn, new_fields_ids_map.as_fields_ids_map())?;
     if let Some(new_primary_key) = new_primary_key {
@@ -132,6 +134,7 @@ pub(super) fn update_index(
     index.put_field_distribution(wtxn, &field_distribution)?;
     index.put_documents_ids(wtxn, &document_ids)?;
     index.set_updated_at(wtxn, &OffsetDateTime::now_utc())?;
+    index.update_documents_stats(wtxn, modified_docids)?;
     Ok(())
 }
 

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

@@ -1,7 +1,9 @@
 mod v1_12;
+mod v1_13;
 
 use heed::RwTxn;
-use v1_12::{V1_12_3_To_Current, V1_12_To_V1_12_3};
+use v1_12::{V1_12_3_To_V1_13_0, V1_12_To_V1_12_3};
+use v1_13::{V1_13_0_To_V1_13_1, V1_13_1_To_Current};
 
 use crate::progress::{Progress, VariableNameStep};
 use crate::{Index, InternalError, Result};
@@ -26,13 +28,19 @@ pub fn upgrade(
     progress: Progress,
 ) -> Result<bool> {
     let from = index.get_version(wtxn)?.unwrap_or(db_version);
-    let upgrade_functions: &[&dyn UpgradeIndex] = &[&V1_12_To_V1_12_3 {}, &V1_12_3_To_Current()];
+    let upgrade_functions: &[&dyn UpgradeIndex] = &[
+        &V1_12_To_V1_12_3 {},
+        &V1_12_3_To_V1_13_0 {},
+        &V1_13_0_To_V1_13_1 {},
+        &V1_13_1_To_Current {},
+    ];
 
     let start = match from {
         (1, 12, 0..=2) => 0,
         (1, 12, 3..) => 1,
+        (1, 13, 0) => 2,
         // We must handle the current version in the match because in case of a failure some index may have been upgraded but not other.
-        (1, 13, _) => return Ok(false),
+        (1, 13, _) => 3,
         (major, minor, patch) => {
             return Err(InternalError::CannotUpgradeToVersion(major, minor, patch).into())
         }

crates/milli/src/update/upgrade/v1_12.rs

@@ -1,7 +1,6 @@
 use heed::RwTxn;
 
 use super::UpgradeIndex;
-use crate::constants::{VERSION_MAJOR, VERSION_MINOR, VERSION_PATCH};
 use crate::progress::Progress;
 use crate::{make_enum_progress, Index, Result};
 
@@ -32,9 +31,9 @@ impl UpgradeIndex for V1_12_To_V1_12_3 {
 }
 
 #[allow(non_camel_case_types)]
-pub(super) struct V1_12_3_To_Current();
+pub(super) struct V1_12_3_To_V1_13_0 {}
 
-impl UpgradeIndex for V1_12_3_To_Current {
+impl UpgradeIndex for V1_12_3_To_V1_13_0 {
     fn upgrade(
         &self,
         _wtxn: &mut RwTxn,
@@ -42,14 +41,11 @@ impl UpgradeIndex for V1_12_3_To_Current {
         _original: (u32, u32, u32),
         _progress: Progress,
     ) -> Result<bool> {
-        Ok(false)
+        // recompute the indexes stats
+        Ok(true)
     }
 
     fn target_version(&self) -> (u32, u32, u32) {
-        (
-            VERSION_MAJOR.parse().unwrap(),
-            VERSION_MINOR.parse().unwrap(),
-            VERSION_PATCH.parse().unwrap(),
-        )
+        (1, 13, 0)
     }
 }

crates/milli/src/update/upgrade/v1_13.rs

@@ -0,0 +1,60 @@
+use heed::RwTxn;
+
+use super::UpgradeIndex;
+use crate::constants::{VERSION_MAJOR, VERSION_MINOR, VERSION_PATCH};
+use crate::database_stats::DatabaseStats;
+use crate::progress::Progress;
+use crate::{make_enum_progress, Index, Result};
+
+#[allow(non_camel_case_types)]
+pub(super) struct V1_13_0_To_V1_13_1();
+
+impl UpgradeIndex for V1_13_0_To_V1_13_1 {
+    fn upgrade(
+        &self,
+        wtxn: &mut RwTxn,
+        index: &Index,
+        _original: (u32, u32, u32),
+        progress: Progress,
+    ) -> Result<bool> {
+        make_enum_progress! {
+            enum DocumentsStats {
+                CreatingDocumentsStats,
+            }
+        };
+
+        // Create the new documents stats.
+        progress.update_progress(DocumentsStats::CreatingDocumentsStats);
+        let stats = DatabaseStats::new(index.documents.remap_types(), wtxn)?;
+        index.put_documents_stats(wtxn, stats)?;
+
+        Ok(true)
+    }
+
+    fn target_version(&self) -> (u32, u32, u32) {
+        (1, 13, 1)
+    }
+}
+
+#[allow(non_camel_case_types)]
+pub(super) struct V1_13_1_To_Current();
+
+impl UpgradeIndex for V1_13_1_To_Current {
+    fn upgrade(
+        &self,
+        _wtxn: &mut RwTxn,
+        _index: &Index,
+        _original: (u32, u32, u32),
+        _progress: Progress,
+    ) -> Result<bool> {
+        Ok(false)
+    }
+
+    fn target_version(&self) -> (u32, u32, u32) {
+        (
+            VERSION_MAJOR.parse().unwrap(),
+            VERSION_MINOR.parse().unwrap(),
+            VERSION_PATCH.parse().unwrap(),
+        )
+    }
+}

crates/milli/src/vector/mod.rs

@@ -410,8 +410,43 @@ impl ArroyWrapper {
     fn quantized_db(&self) -> arroy::Database<BinaryQuantizedCosine> {
         self.database.remap_data_type()
     }
+
+    pub fn aggregate_stats(
+        &self,
+        rtxn: &RoTxn,
+        stats: &mut ArroyStats,
+    ) -> Result<(), arroy::Error> {
+        if self.quantized {
+            for reader in self.readers(rtxn, self.quantized_db()) {
+                let reader = reader?;
+                let documents = reader.item_ids();
+                if documents.is_empty() {
+                    break;
+                }
+                stats.documents |= documents;
+                stats.number_of_embeddings += documents.len();
+            }
+        } else {
+            for reader in self.readers(rtxn, self.angular_db()) {
+                let reader = reader?;
+                let documents = reader.item_ids();
+                if documents.is_empty() {
+                    break;
+                }
+                stats.documents |= documents;
+                stats.number_of_embeddings += documents.len();
+            }
+        }
+
+        Ok(())
+    }
 }
 
+#[derive(Debug, Default, Clone)]
+pub struct ArroyStats {
+    pub number_of_embeddings: u64,
+    pub documents: RoaringBitmap,
+}
 /// One or multiple embeddings stored consecutively in a flat vector.
 pub struct Embeddings<F> {
     data: Vec<F>,

crates/milli/src/vector/rest.rs

@@ -130,6 +130,7 @@ impl Embedder {
         let client = ureq::AgentBuilder::new()
             .max_idle_connections(REQUEST_PARALLELISM * 2)
             .max_idle_connections_per_host(REQUEST_PARALLELISM * 2)
+            .timeout(std::time::Duration::from_secs(30))
             .build();
 
         let request = Request::new(options.request)?;

workloads/hackernews-modify-facet-numbers.json

@@ -31,7 +31,7 @@
     "hackernews-modified-number-filters.ndjson": {
       "local_location": null,
       "remote_location": "https://milli-benchmarks.fra1.digitaloceanspaces.com/bench/datasets/hackernews/modification/01-modified-filters.ndjson",
-      "sha256": "7272cbfd41110d32d7fe168424a0000f07589bfe40f664652b34f4f20aaf3802"
+      "sha256": "b80c245ce1b1df80b9b38800f677f3bd11947ebc62716fb108269d50e796c35c"
     }
   },
   "precommands": [

workloads/hackernews-modify-facet-strings.json

@@ -31,7 +31,7 @@
     "hackernews-modified-string-filters.ndjson": {
       "local_location": null,
       "remote_location": "https://milli-benchmarks.fra1.digitaloceanspaces.com/bench/datasets/hackernews/modification/02-modified-filters.ndjson",
-      "sha256": "b80c245ce1b1df80b9b38800f677f3bd11947ebc62716fb108269d50e796c35c"
+      "sha256": "7272cbfd41110d32d7fe168424a0000f07589bfe40f664652b34f4f20aaf3802"
     }
   },
   "precommands": [