Move MostlySend, ThreadLocal, FullySend to their own commit

crates/index-scheduler/src/batch.rs

@@ -32,18 +32,15 @@ use meilisearch_types::error::Code;
 use meilisearch_types::heed::{RoTxn, RwTxn};
 use meilisearch_types::milli::documents::{obkv_to_object, DocumentsBatchReader, PrimaryKey};
 use meilisearch_types::milli::heed::CompactionOption;
-use meilisearch_types::milli::update::new::indexer::{
-    self, retrieve_or_guess_primary_key, UpdateByFunction,
-};
+use meilisearch_types::milli::update::new::indexer::{self, UpdateByFunction};
 use meilisearch_types::milli::update::{IndexDocumentsMethod, Settings as MilliSettings};
 use meilisearch_types::milli::vector::parsed_vectors::{
     ExplicitVectors, VectorOrArrayOfVectors, RESERVED_VECTORS_FIELD_NAME,
 };
-use meilisearch_types::milli::{self, Filter, ThreadPoolNoAbort, ThreadPoolNoAbortBuilder};
+use meilisearch_types::milli::{self, Filter, ThreadPoolNoAbortBuilder};
 use meilisearch_types::settings::{apply_settings_to_builder, Settings, Unchecked};
 use meilisearch_types::tasks::{Details, IndexSwap, Kind, KindWithContent, Status, Task};
 use meilisearch_types::{compression, Index, VERSION_FILE_NAME};
-use raw_collections::RawMap;
 use roaring::RoaringBitmap;
 use time::macros::format_description;
 use time::OffsetDateTime;
@@ -1229,9 +1226,7 @@ impl IndexScheduler {
         const PRINT_SECS_DELTA: u64 = 1;
 
         let processing_tasks = self.processing_tasks.clone();
-
         let must_stop_processing = self.must_stop_processing.clone();
-
         let send_progress = |progress| {
             let now = std::time::Instant::now();
             let elapsed = secs_since_started_processing_at.load(atomic::Ordering::Relaxed);
@@ -1280,16 +1275,6 @@ impl IndexScheduler {
                 // TODO: at some point, for better efficiency we might want to reuse the bumpalo for successive batches.
                 // this is made difficult by the fact we're doing private clones of the index scheduler and sending it
                 // to a fresh thread.
-
-                /// TODO manage errors correctly
-                let first_addition_uuid = operations
-                    .iter()
-                    .find_map(|op| match op {
-                        DocumentOperation::Add(content_uuid) => Some(content_uuid),
-                        _ => None,
-                    })
-                    .unwrap();
-
                 let mut content_files = Vec::new();
                 for operation in &operations {
                     if let DocumentOperation::Add(content_uuid) = operation {
@@ -1305,88 +1290,30 @@ impl IndexScheduler {
                 let db_fields_ids_map = index.fields_ids_map(&rtxn)?;
                 let mut new_fields_ids_map = db_fields_ids_map.clone();
 
-                let first_document = match content_files.first() {
-                    Some(mmap) => {
-                        let mut iter = serde_json::Deserializer::from_slice(mmap).into_iter();
-                        iter.next().transpose().map_err(|e| e.into()).map_err(Error::IoError)?
-                    }
-                    None => None,
-                };
-
-                let (primary_key, primary_key_has_been_set) = retrieve_or_guess_primary_key(
-                    &rtxn,
-                    index,
-                    &mut new_fields_ids_map,
-                    primary_key.as_deref(),
-                    first_document
-                        .map(|raw| RawMap::from_raw_value(raw, &indexer_alloc))
-                        .transpose()
-                        .map_err(|error| {
-                            milli::Error::UserError(milli::UserError::SerdeJson(error))
-                        })?,
-                )?
-                .map_err(milli::Error::from)?;
-
                 let mut content_files_iter = content_files.iter();
                 let mut indexer = indexer::DocumentOperation::new(method);
                 let embedders = index.embedding_configs(index_wtxn)?;
                 let embedders = self.embedders(embedders)?;
-                for (operation, task) in operations.into_iter().zip(tasks.iter_mut()) {
+                for operation in operations {
                     match operation {
                         DocumentOperation::Add(_content_uuid) => {
                             let mmap = content_files_iter.next().unwrap();
-                            let stats = indexer.add_documents(mmap)?;
+                            indexer.add_documents(mmap)?;
                             // builder = builder.with_embedders(embedders.clone());
-
-                            let received_documents =
-                                if let Some(Details::DocumentAdditionOrUpdate {
-                                    received_documents,
-                                    ..
-                                }) = task.details
-                                {
-                                    received_documents
-                                } else {
-                                    // In the case of a `documentAdditionOrUpdate` the details MUST be set
-                                    unreachable!();
-                                };
-
-                            task.status = Status::Succeeded;
-                            task.details = Some(Details::DocumentAdditionOrUpdate {
-                                received_documents,
-                                indexed_documents: Some(stats.document_count as u64),
-                            })
                         }
                         DocumentOperation::Delete(document_ids) => {
-                            let count = document_ids.len();
                             let document_ids: bumpalo::collections::vec::Vec<_> = document_ids
                                 .iter()
                                 .map(|s| &*indexer_alloc.alloc_str(s))
                                 .collect_in(&indexer_alloc);
                             indexer.delete_documents(document_ids.into_bump_slice());
-                            // Uses Invariant: remove documents actually always returns Ok for the inner result
-                            // let count = user_result.unwrap();
-                            let provided_ids =
-                                if let Some(Details::DocumentDeletion { provided_ids, .. }) =
-                                    task.details
-                                {
-                                    provided_ids
-                                } else {
-                                    // In the case of a `documentAdditionOrUpdate` the details MUST be set
-                                    unreachable!();
-                                };
-
-                            task.status = Status::Succeeded;
-                            task.details = Some(Details::DocumentDeletion {
-                                provided_ids,
-                                deleted_documents: Some(count as u64),
-                            });
                         }
                     }
                 }
 
-                if tasks.iter().any(|res| res.error.is_none()) {
                 let local_pool;
-                    let pool = match &self.index_mapper.indexer_config().thread_pool {
+                let indexer_config = self.index_mapper.indexer_config();
+                let pool = match &indexer_config.thread_pool {
                     Some(pool) => pool,
                     None => {
                         local_pool = ThreadPoolNoAbortBuilder::new().build().unwrap();
@@ -1394,22 +1321,55 @@ impl IndexScheduler {
                     }
                 };
 
-                    // TODO we want to multithread this
-                    let document_changes = indexer.into_changes(
+                let (document_changes, operation_stats, primary_key) = indexer.into_changes(
                     &indexer_alloc,
                     index,
                     &rtxn,
-                        &primary_key,
+                    primary_key.as_deref(),
                     &mut new_fields_ids_map,
                 )?;
 
+                let mut addition = 0;
+                for (stats, task) in operation_stats.into_iter().zip(&mut tasks) {
+                    addition += stats.document_count;
+                    match stats.error {
+                        Some(error) => {
+                            task.status = Status::Failed;
+                            task.error = Some(milli::Error::UserError(error).into());
+                        }
+                        None => task.status = Status::Succeeded,
+                    }
+
+                    task.details = match task.details {
+                        Some(Details::DocumentAdditionOrUpdate { received_documents, .. }) => {
+                            Some(Details::DocumentAdditionOrUpdate {
+                                received_documents,
+                                indexed_documents: Some(stats.document_count),
+                            })
+                        }
+                        Some(Details::DocumentDeletion { provided_ids, .. }) => {
+                            Some(Details::DocumentDeletion {
+                                provided_ids,
+                                deleted_documents: Some(stats.document_count),
+                            })
+                        }
+                        _ => {
+                            // In the case of a `documentAdditionOrUpdate` or `DocumentDeletion`
+                            // the details MUST be set to either addition or deletion
+                            unreachable!();
+                        }
+                    }
+                }
+
+                if tasks.iter().any(|res| res.error.is_none()) {
                     pool.install(|| {
                         indexer::index(
                             index_wtxn,
                             index,
+                            indexer_config.grenad_parameters(),
                             &db_fields_ids_map,
                             new_fields_ids_map,
-                            primary_key_has_been_set.then_some(primary_key),
+                            primary_key,
                             &document_changes,
                             embedders,
                             &|| must_stop_processing.get(),
@@ -1418,7 +1378,7 @@ impl IndexScheduler {
                     })
                     .unwrap()?;
 
-                    // tracing::info!(indexing_result = ?addition, processed_in = ?started_processing_at.elapsed(), "document indexing done");
+                    tracing::info!(indexing_result = ?addition, processed_in = ?started_processing_at.elapsed(), "document indexing done");
                 }
                 // else if primary_key_has_been_set {
                 //     // Everything failed but we've set a primary key.
@@ -1435,12 +1395,14 @@ impl IndexScheduler {
                 Ok(tasks)
             }
             IndexOperation::DocumentEdition { mut task, .. } => {
-                let (filter, context, code) =
-                    if let KindWithContent::DocumentEdition {
-                        filter_expr, context, function, ..
+                let (filter, code) = if let KindWithContent::DocumentEdition {
+                    filter_expr,
+                    context: _,
+                    function,
+                    ..
                 } = &task.kind
                 {
-                        (filter_expr, context, function)
+                    (filter_expr, function)
                 } else {
                     unreachable!()
                 };
@@ -1497,7 +1459,8 @@ impl IndexScheduler {
 
                 if task.error.is_none() {
                     let local_pool;
-                    let pool = match &self.index_mapper.indexer_config().thread_pool {
+                    let indexer_config = self.index_mapper.indexer_config();
+                    let pool = match &indexer_config.thread_pool {
                         Some(pool) => pool,
                         None => {
                             local_pool = ThreadPoolNoAbortBuilder::new().build().unwrap();
@@ -1515,6 +1478,7 @@ impl IndexScheduler {
                         indexer::index(
                             index_wtxn,
                             index,
+                            indexer_config.grenad_parameters(),
                             &db_fields_ids_map,
                             new_fields_ids_map,
                             None, // cannot change primary key in DocumentEdition
@@ -1647,7 +1611,8 @@ impl IndexScheduler {
 
                 if !tasks.iter().all(|res| res.error.is_some()) {
                     let local_pool;
-                    let pool = match &self.index_mapper.indexer_config().thread_pool {
+                    let indexer_config = self.index_mapper.indexer_config();
+                    let pool = match &indexer_config.thread_pool {
                         Some(pool) => pool,
                         None => {
                             local_pool = ThreadPoolNoAbortBuilder::new().build().unwrap();
@@ -1665,6 +1630,7 @@ impl IndexScheduler {
                         indexer::index(
                             index_wtxn,
                             index,
+                            indexer_config.grenad_parameters(),
                             &db_fields_ids_map,
                             new_fields_ids_map,
                             None, // document deletion never changes primary key

crates/index-scheduler/src/lib.rs

@@ -971,6 +971,8 @@ impl IndexScheduler {
         let ProcessingTasks { started_at, processing, progress, .. } =
             self.processing_tasks.read().map_err(|_| Error::CorruptedTaskQueue)?.clone();
 
+        let _ = progress;
+
         let ret = tasks.into_iter();
         if processing.is_empty() {
             Ok((ret.collect(), total))
@@ -4296,11 +4298,11 @@ mod tests {
         snapshot!(snapshot_index_scheduler(&index_scheduler), name: "only_first_task_succeed");
 
         // The second batch should fail.
-        handle.advance_one_failed_batch();
+        handle.advance_one_successful_batch();
         snapshot!(snapshot_index_scheduler(&index_scheduler), name: "second_task_fails");
 
         // The second batch should fail.
-        handle.advance_one_failed_batch();
+        handle.advance_one_successful_batch();
         snapshot!(snapshot_index_scheduler(&index_scheduler), name: "third_task_fails");
 
         // Is the primary key still what we expect?
@@ -4361,7 +4363,7 @@ mod tests {
         snapshot!(snapshot_index_scheduler(&index_scheduler), name: "only_first_task_succeed");
 
         // The second batch should fail and contains two tasks.
-        handle.advance_one_failed_batch();
+        handle.advance_one_successful_batch();
         snapshot!(snapshot_index_scheduler(&index_scheduler), name: "second_and_third_tasks_fails");
 
         // Is the primary key still what we expect?
@@ -4440,7 +4442,8 @@ mod tests {
         snapshot!(primary_key, @"id");
 
         // We're trying to `bork` again, but now there is already a primary key set for this index.
-        handle.advance_one_failed_batch();
+        // NOTE: it's marked as successful because the batch didn't fails, it's the individual tasks that failed.
+        handle.advance_one_successful_batch();
         snapshot!(snapshot_index_scheduler(&index_scheduler), name: "fourth_task_fails");
 
         // Finally the last task should succeed since its primary key is the same as the valid one.
@@ -4600,7 +4603,7 @@ mod tests {
         snapshot!(primary_key.is_none(), @"false");
 
         // The second batch should contains only one task that fails because it tries to update the primary key to `bork`.
-        handle.advance_one_failed_batch();
+        handle.advance_one_successful_batch();
         snapshot!(snapshot_index_scheduler(&index_scheduler), name: "second_task_fails");
 
         // The third batch should succeed and only contains one task.
@@ -5213,9 +5216,10 @@ mod tests {
 
             let configs = index_scheduler.embedders(configs).unwrap();
             let (hf_embedder, _, _) = configs.get(&simple_hf_name).unwrap();
-            let beagle_embed = hf_embedder.embed_one(S("Intel the beagle best doggo")).unwrap();
-            let lab_embed = hf_embedder.embed_one(S("Max the lab best doggo")).unwrap();
-            let patou_embed = hf_embedder.embed_one(S("kefir the patou best doggo")).unwrap();
+            let beagle_embed =
+                hf_embedder.embed_one(S("Intel the beagle best doggo"), None).unwrap();
+            let lab_embed = hf_embedder.embed_one(S("Max the lab best doggo"), None).unwrap();
+            let patou_embed = hf_embedder.embed_one(S("kefir the patou best doggo"), None).unwrap();
             (fakerest_name, simple_hf_name, beagle_embed, lab_embed, patou_embed)
         };
 

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

@@ -1,5 +1,5 @@
 ---
-source: index-scheduler/src/lib.rs
+source: crates/index-scheduler/src/lib.rs
 ---
 ### Autobatching Enabled = true
 ### Processing Tasks:
@@ -22,7 +22,7 @@ succeeded [0,1,2,]
 doggos [0,1,2,]
 ----------------------------------------------------------------------
 ### Index Mapper:
-doggos: { number_of_documents: 1, field_distribution: {"_vectors": 1, "breed": 1, "doggo": 1, "id": 1} }
+doggos: { number_of_documents: 1, field_distribution: {"breed": 1, "doggo": 1, "id": 1} }
 
 ----------------------------------------------------------------------
 ### Canceled By:

crates/meilisearch-types/src/document_formats.rs

@@ -220,11 +220,12 @@ pub fn read_json(input: &File, output: impl io::Write) -> Result<u64> {
 
     let mut out = BufWriter::new(output);
     let mut deserializer = serde_json::Deserializer::from_slice(&input);
-    let count = match array_each(&mut deserializer, |obj: &RawValue| {
+    let res = array_each(&mut deserializer, |obj: &RawValue| {
         doc_alloc.reset();
         let map = RawMap::from_raw_value(obj, &doc_alloc)?;
         to_writer(&mut out, &map)
-    }) {
+    });
+    let count = match res {
         // The json data has been deserialized and does not need to be processed again.
         // The data has been transferred to the writer during the deserialization process.
         Ok(Ok(count)) => count,

crates/meilisearch/src/search/mod.rs

@@ -796,8 +796,10 @@ fn prepare_search<'t>(
                     let span = tracing::trace_span!(target: "search::vector", "embed_one");
                     let _entered = span.enter();
 
+                    let deadline = std::time::Instant::now() + std::time::Duration::from_secs(10);
+
                     embedder
-                        .embed_one(query.q.clone().unwrap())
+                        .embed_one(query.q.clone().unwrap(), Some(deadline))
                         .map_err(milli::vector::Error::from)
                         .map_err(milli::Error::from)?
                 }

crates/meilisearch/tests/search/multi.rs

@@ -585,9 +585,9 @@ async fn federation_two_search_two_indexes() {
           ],
           "_vectors": {
             "manual": [
-              -100.0,
-              340.0,
-              90.0
+              -100,
+              340,
+              90
             ]
           },
           "_federation": {
@@ -613,9 +613,9 @@ async fn federation_two_search_two_indexes() {
           "cattos": "pésti",
           "_vectors": {
             "manual": [
-              1.0,
-              2.0,
-              3.0
+              1,
+              2,
+              3
             ]
           },
           "_federation": {
@@ -640,9 +640,9 @@ async fn federation_two_search_two_indexes() {
           ],
           "_vectors": {
             "manual": [
-              1.0,
-              2.0,
-              54.0
+              1,
+              2,
+              54
             ]
           },
           "_federation": {
@@ -707,9 +707,9 @@ async fn federation_multiple_search_multiple_indexes() {
           ],
           "_vectors": {
             "manual": [
-              -100.0,
-              340.0,
-              90.0
+              -100,
+              340,
+              90
             ]
           },
           "_federation": {
@@ -735,9 +735,9 @@ async fn federation_multiple_search_multiple_indexes() {
           "cattos": "pésti",
           "_vectors": {
             "manual": [
-              1.0,
-              2.0,
-              3.0
+              1,
+              2,
+              3
             ]
           },
           "_federation": {
@@ -773,9 +773,9 @@ async fn federation_multiple_search_multiple_indexes() {
           ],
           "_vectors": {
             "manual": [
-              1.0,
-              2.0,
-              54.0
+              1,
+              2,
+              54
             ]
           },
           "_federation": {
@@ -793,9 +793,9 @@ async fn federation_multiple_search_multiple_indexes() {
           ],
           "_vectors": {
             "manual": [
-              10.0,
-              -23.0,
-              32.0
+              10,
+              -23,
+              32
             ]
           },
           "_federation": {
@@ -824,9 +824,9 @@ async fn federation_multiple_search_multiple_indexes() {
           ],
           "_vectors": {
             "manual": [
-              10.0,
-              23.0,
-              32.0
+              10,
+              23,
+              32
             ]
           },
           "_federation": {
@@ -869,9 +869,9 @@ async fn federation_multiple_search_multiple_indexes() {
           ],
           "_vectors": {
             "manual": [
-              1.0,
-              2.0,
-              54.0
+              1,
+              2,
+              54
             ]
           },
           "_federation": {
@@ -898,9 +898,9 @@ async fn federation_multiple_search_multiple_indexes() {
           ],
           "_vectors": {
             "manual": [
-              -100.0,
-              231.0,
-              32.0
+              -100,
+              231,
+              32
             ]
           },
           "_federation": {
@@ -1522,9 +1522,9 @@ async fn federation_sort_same_indexes_same_criterion_same_direction() {
           "cattos": "pésti",
           "_vectors": {
             "manual": [
-              1.0,
-              2.0,
-              3.0
+              1,
+              2,
+              3
             ]
           },
           "_federation": {
@@ -1550,9 +1550,9 @@ async fn federation_sort_same_indexes_same_criterion_same_direction() {
           ],
           "_vectors": {
             "manual": [
-              1.0,
-              2.0,
-              54.0
+              1,
+              2,
+              54
             ]
           },
           "_federation": {
@@ -1582,9 +1582,9 @@ async fn federation_sort_same_indexes_same_criterion_same_direction() {
           ],
           "_vectors": {
             "manual": [
-              10.0,
-              23.0,
-              32.0
+              10,
+              23,
+              32
             ]
           },
           "_federation": {
@@ -1845,9 +1845,9 @@ async fn federation_sort_same_indexes_different_criterion_same_direction() {
           ],
           "_vectors": {
             "manual": [
-              1.0,
-              2.0,
-              54.0
+              1,
+              2,
+              54
             ]
           },
           "_federation": {
@@ -1874,9 +1874,9 @@ async fn federation_sort_same_indexes_different_criterion_same_direction() {
           "cattos": "pésti",
           "_vectors": {
             "manual": [
-              1.0,
-              2.0,
-              3.0
+              1,
+              2,
+              3
             ]
           },
           "_federation": {
@@ -1906,9 +1906,9 @@ async fn federation_sort_same_indexes_different_criterion_same_direction() {
           ],
           "_vectors": {
             "manual": [
-              10.0,
-              23.0,
-              32.0
+              10,
+              23,
+              32
             ]
           },
           "_federation": {

crates/meilisearch/tests/vector/openai.rs

@@ -137,13 +137,14 @@ fn long_text() -> &'static str {
 }
 
 async fn create_mock_tokenized() -> (MockServer, Value) {
-    create_mock_with_template("{{doc.text}}", ModelDimensions::Large, false).await
+    create_mock_with_template("{{doc.text}}", ModelDimensions::Large, false, false).await
 }
 
 async fn create_mock_with_template(
     document_template: &str,
     model_dimensions: ModelDimensions,
     fallible: bool,
+    slow: bool,
 ) -> (MockServer, Value) {
     let mock_server = MockServer::start().await;
     const API_KEY: &str = "my-api-key";
@@ -154,7 +155,11 @@ async fn create_mock_with_template(
     Mock::given(method("POST"))
         .and(path("/"))
         .respond_with(move |req: &Request| {
-            // 0. maybe return 500
+            // 0. wait for a long time
+            if slow {
+              std::thread::sleep(std::time::Duration::from_secs(1));
+            }
+            // 1. maybe return 500
             if fallible {
              let attempt = attempt.fetch_add(1, Ordering::Relaxed);
              let failed = matches!(attempt % 4, 0 | 1 | 3);
@@ -167,7 +172,7 @@ async fn create_mock_with_template(
                 }))
              }
             }
-            // 1. check API key
+            // 3. check API key
             match req.headers.get("Authorization") {
                 Some(api_key) if api_key == API_KEY_BEARER => {
                     {}
@@ -202,7 +207,7 @@ async fn create_mock_with_template(
                     )
                 }
             }
-            // 2. parse text inputs
+            // 3. parse text inputs
             let query: serde_json::Value = match req.body_json() {
                 Ok(query) => query,
                 Err(_error) => return ResponseTemplate::new(400).set_body_json(
@@ -223,7 +228,7 @@ async fn create_mock_with_template(
                 panic!("Expected {model_dimensions:?}, got {query_model_dimensions:?}")
             }
 
-            // 3. for each text, find embedding in responses
+            // 4. for each text, find embedding in responses
             let serde_json::Value::Array(inputs) = &query["input"] else {
                 panic!("Unexpected `input` value")
             };
@@ -283,7 +288,7 @@ async fn create_mock_with_template(
                 "embedding": embedding,
             })).collect();
 
-            // 4. produce output from embeddings
+            // 5. produce output from embeddings
             ResponseTemplate::new(200).set_body_json(json!({
                 "object": "list",
                 "data": data,
@@ -317,23 +322,27 @@ const DOGGO_TEMPLATE: &str = r#"{%- if doc.gender == "F" -%}Une chienne nommée
         {%- endif %}, de race {{doc.breed}}."#;
 
 async fn create_mock() -> (MockServer, Value) {
-    create_mock_with_template(DOGGO_TEMPLATE, ModelDimensions::Large, false).await
+    create_mock_with_template(DOGGO_TEMPLATE, ModelDimensions::Large, false, false).await
 }
 
 async fn create_mock_dimensions() -> (MockServer, Value) {
-    create_mock_with_template(DOGGO_TEMPLATE, ModelDimensions::Large512, false).await
+    create_mock_with_template(DOGGO_TEMPLATE, ModelDimensions::Large512, false, false).await
 }
 
 async fn create_mock_small_embedding_model() -> (MockServer, Value) {
-    create_mock_with_template(DOGGO_TEMPLATE, ModelDimensions::Small, false).await
+    create_mock_with_template(DOGGO_TEMPLATE, ModelDimensions::Small, false, false).await
 }
 
 async fn create_mock_legacy_embedding_model() -> (MockServer, Value) {
-    create_mock_with_template(DOGGO_TEMPLATE, ModelDimensions::Ada, false).await
+    create_mock_with_template(DOGGO_TEMPLATE, ModelDimensions::Ada, false, false).await
 }
 
 async fn create_fallible_mock() -> (MockServer, Value) {
-    create_mock_with_template(DOGGO_TEMPLATE, ModelDimensions::Large, true).await
+    create_mock_with_template(DOGGO_TEMPLATE, ModelDimensions::Large, true, false).await
+}
+
+async fn create_slow_mock() -> (MockServer, Value) {
+    create_mock_with_template(DOGGO_TEMPLATE, ModelDimensions::Large, true, true).await
 }
 
 // basic test "it works"
@@ -1873,4 +1882,114 @@ async fn it_still_works() {
     ]
     "###);
 }
+
+// test with a server that responds 500 on 3 out of 4 calls
+#[actix_rt::test]
+async fn timeout() {
+    let (_mock, setting) = create_slow_mock().await;
+    let server = get_server_vector().await;
+    let index = server.index("doggo");
+
+    let (response, code) = index
+        .update_settings(json!({
+          "embedders": {
+              "default": setting,
+          },
+        }))
+        .await;
+    snapshot!(code, @"202 Accepted");
+    let task = server.wait_task(response.uid()).await;
+    snapshot!(task["status"], @r###""succeeded""###);
+    let documents = json!([
+      {"id": 0, "name": "kefir", "gender": "M", "birthyear": 2023, "breed": "Patou"},
+    ]);
+    let (value, code) = index.add_documents(documents, None).await;
+    snapshot!(code, @"202 Accepted");
+    let task = index.wait_task(value.uid()).await;
+    snapshot!(task, @r###"
+    {
+      "uid": "[uid]",
+      "indexUid": "doggo",
+      "status": "succeeded",
+      "type": "documentAdditionOrUpdate",
+      "canceledBy": null,
+      "details": {
+        "receivedDocuments": 1,
+        "indexedDocuments": 1
+      },
+      "error": null,
+      "duration": "[duration]",
+      "enqueuedAt": "[date]",
+      "startedAt": "[date]",
+      "finishedAt": "[date]"
+    }
+    "###);
+
+    let (documents, _code) = index
+        .get_all_documents(GetAllDocumentsOptions { retrieve_vectors: true, ..Default::default() })
+        .await;
+    snapshot!(json_string!(documents, {".results.*._vectors.default.embeddings" => "[vector]"}), @r###"
+    {
+      "results": [
+        {
+          "id": 0,
+          "name": "kefir",
+          "gender": "M",
+          "birthyear": 2023,
+          "breed": "Patou",
+          "_vectors": {
+            "default": {
+              "embeddings": "[vector]",
+              "regenerate": true
+            }
+          }
+        }
+      ],
+      "offset": 0,
+      "limit": 20,
+      "total": 1
+    }
+    "###);
+
+    let (response, code) = index
+        .search_post(json!({
+            "q": "grand chien de berger des montagnes",
+            "hybrid": {"semanticRatio": 0.99, "embedder": "default"}
+        }))
+        .await;
+    snapshot!(code, @"200 OK");
+    snapshot!(json_string!(response["semanticHitCount"]), @"0");
+    snapshot!(json_string!(response["hits"]), @"[]");
+
+    let (response, code) = index
+        .search_post(json!({
+            "q": "grand chien de berger des montagnes",
+            "hybrid": {"semanticRatio": 0.99, "embedder": "default"}
+        }))
+        .await;
+    snapshot!(code, @"200 OK");
+    snapshot!(json_string!(response["semanticHitCount"]), @"1");
+    snapshot!(json_string!(response["hits"]), @r###"
+    [
+      {
+        "id": 0,
+        "name": "kefir",
+        "gender": "M",
+        "birthyear": 2023,
+        "breed": "Patou"
+      }
+    ]
+    "###);
+
+    let (response, code) = index
+        .search_post(json!({
+            "q": "grand chien de berger des montagnes",
+            "hybrid": {"semanticRatio": 0.99, "embedder": "default"}
+        }))
+        .await;
+    snapshot!(code, @"200 OK");
+    snapshot!(json_string!(response["semanticHitCount"]), @"0");
+    snapshot!(json_string!(response["hits"]), @"[]");
+}
+
 // test with a server that wrongly responds 400

crates/milli/src/search/hybrid.rs

@@ -201,7 +201,9 @@ impl<'a> Search<'a> {
                 let span = tracing::trace_span!(target: "search::hybrid", "embed_one");
                 let _entered = span.enter();
 
-                match embedder.embed_one(query) {
+                let deadline = std::time::Instant::now() + std::time::Duration::from_secs(3);
+
+                match embedder.embed_one(query, Some(deadline)) {
                     Ok(embedding) => embedding,
                     Err(error) => {
                         tracing::error!(error=%error, "Embedding failed");

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

@@ -119,12 +119,8 @@ impl GrenadParameters {
     ///
     /// This should be called inside of a rayon thread pool,
     /// otherwise, it will take the global number of threads.
-    ///
-    /// The max memory cannot exceed a given reasonable value.
     pub fn max_memory_by_thread(&self) -> Option<usize> {
-        self.max_memory.map(|max_memory| {
-            (max_memory / rayon::current_num_threads()).min(MAX_GRENAD_SORTER_USAGE)
-        })
+        self.max_memory.map(|max_memory| (max_memory / rayon::current_num_threads()))
     }
 }
 

crates/milli/src/update/indexer_config.rs

@@ -1,5 +1,6 @@
 use grenad::CompressionType;
 
+use super::GrenadParameters;
 use crate::thread_pool_no_abort::ThreadPoolNoAbort;
 
 #[derive(Debug)]
@@ -15,6 +16,17 @@ pub struct IndexerConfig {
     pub skip_index_budget: bool,
 }
 
+impl IndexerConfig {
+    pub fn grenad_parameters(&self) -> GrenadParameters {
+        GrenadParameters {
+            chunk_compression_type: self.chunk_compression_type,
+            chunk_compression_level: self.chunk_compression_level,
+            max_memory: self.max_memory,
+            max_nb_chunks: self.max_nb_chunks,
+        }
+    }
+}
+
 impl Default for IndexerConfig {
     fn default() -> Self {
         Self {

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

@@ -70,28 +70,30 @@ impl<'t, Mapper: FieldIdMapper> Document<'t> for DocumentFromDb<'t, Mapper> {
     fn iter_top_level_fields(&self) -> impl Iterator<Item = Result<(&'t str, &'t RawValue)>> {
         let mut it = self.content.iter();
 
-        std::iter::from_fn(move || {
+        std::iter::from_fn(move || loop {
             let (fid, value) = it.next()?;
-
-            let res = (|| loop {
-                let name = self.fields_ids_map.name(fid).ok_or(
+            let name = match self.fields_ids_map.name(fid).ok_or(
                 InternalError::FieldIdMapMissingEntry(crate::FieldIdMapMissingEntry::FieldId {
                     field_id: fid,
                     process: "getting current document",
                 }),
-                )?;
+            ) {
+                Ok(name) => name,
+                Err(error) => return Some(Err(error.into())),
+            };
 
             if name == RESERVED_VECTORS_FIELD_NAME || name == "_geo" {
                 continue;
             }
 
+            let res = (|| {
                 let value =
                     serde_json::from_slice(value).map_err(crate::InternalError::SerdeJson)?;
 
-                return Ok((name, value));
+                Ok((name, value))
             })();
 
-            Some(res)
+            return Some(res);
         })
     }
 

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

@@ -79,7 +79,7 @@ use roaring::RoaringBitmap;
 use rustc_hash::FxBuildHasher;
 
 use crate::update::del_add::{DelAdd, KvWriterDelAdd};
-use crate::update::new::indexer::document_changes::MostlySend;
+use crate::update::new::thread_local::MostlySend;
 use crate::update::new::KvReaderDelAdd;
 use crate::update::MergeDeladdCboRoaringBitmaps;
 use crate::{CboRoaringBitmapCodec, Result};

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

@@ -6,8 +6,9 @@ use hashbrown::HashMap;
 use super::DelAddRoaringBitmap;
 use crate::update::new::channel::DocumentsSender;
 use crate::update::new::document::{write_to_obkv, Document as _};
-use crate::update::new::indexer::document_changes::{DocumentChangeContext, Extractor, FullySend};
+use crate::update::new::indexer::document_changes::{DocumentChangeContext, Extractor};
 use crate::update::new::ref_cell_ext::RefCellExt as _;
+use crate::update::new::thread_local::FullySend;
 use crate::update::new::DocumentChange;
 use crate::vector::EmbeddingConfigs;
 use crate::Result;
@@ -43,10 +44,12 @@ impl<'a, 'extractor> Extractor<'extractor> for DocumentsExtractor<'a> {
         let mut document_buffer = bumpalo::collections::Vec::new_in(&context.doc_alloc);
         let mut document_extractor_data = context.data.0.borrow_mut_or_yield();
 
-        let mut new_fields_ids_map = context.new_fields_ids_map.borrow_mut_or_yield();
-
         for change in changes {
             let change = change?;
+            // **WARNING**: the exclusive borrow on `new_fields_ids_map` needs to be taken **inside** of the `for change in changes` loop
+            // Otherwise, `BorrowMutError` will occur for document changes that also need the new_fields_ids_map (e.g.: UpdateByFunction)
+            let mut new_fields_ids_map = context.new_fields_ids_map.borrow_mut_or_yield();
+
             let external_docid = change.external_docid().to_owned();
 
             // document but we need to create a function that collects and compresses documents.

crates/milli/src/update/new/extract/faceted/extract_facets.rs

@@ -15,10 +15,10 @@ use crate::heed_codec::facet::OrderedF64Codec;
 use crate::update::del_add::DelAdd;
 use crate::update::new::channel::FieldIdDocidFacetSender;
 use crate::update::new::indexer::document_changes::{
-    extract, DocumentChangeContext, DocumentChanges, Extractor, FullySend, IndexingContext,
-    Progress, ThreadLocal,
+    extract, DocumentChangeContext, DocumentChanges, Extractor, IndexingContext, Progress,
 };
 use crate::update::new::ref_cell_ext::RefCellExt as _;
+use crate::update::new::thread_local::{FullySend, ThreadLocal};
 use crate::update::new::DocumentChange;
 use crate::update::GrenadParameters;
 use crate::{DocumentId, FieldId, Index, Result, MAX_FACET_VALUE_LENGTH};
@@ -36,7 +36,7 @@ impl<'a, 'extractor> Extractor<'extractor> for FacetedExtractorData<'a> {
     fn init_data(&self, extractor_alloc: &'extractor Bump) -> Result<Self::Data> {
         Ok(RefCell::new(BalancedCaches::new_in(
             self.buckets,
-            self.grenad_parameters.max_memory,
+            self.grenad_parameters.max_memory_by_thread(),
             extractor_alloc,
         )))
     }
@@ -156,6 +156,7 @@ impl FacetedDocidsExtractor {
         res
     }
 
+    #[allow(clippy::too_many_arguments)]
     fn facet_fn_with_options<'extractor, 'doc>(
         doc_alloc: &'doc Bump,
         cached_sorter: &mut BalancedCaches<'extractor>,
@@ -336,6 +337,7 @@ fn truncate_str(s: &str) -> &str {
 }
 
 impl FacetedDocidsExtractor {
+    #[allow(clippy::too_many_arguments)]
     #[tracing::instrument(level = "trace", skip_all, target = "indexing::extract::faceted")]
     pub fn run_extraction<
         'pl,

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

@@ -11,8 +11,9 @@ use serde_json::Value;
 
 use crate::error::GeoError;
 use crate::update::new::document::Document;
-use crate::update::new::indexer::document_changes::{DocumentChangeContext, Extractor, MostlySend};
+use crate::update::new::indexer::document_changes::{DocumentChangeContext, Extractor};
 use crate::update::new::ref_cell_ext::RefCellExt as _;
+use crate::update::new::thread_local::MostlySend;
 use crate::update::new::DocumentChange;
 use crate::update::GrenadParameters;
 use crate::{lat_lng_to_xyz, DocumentId, GeoPoint, Index, InternalError, Result};
@@ -150,7 +151,7 @@ impl<'extractor> Extractor<'extractor> for GeoExtractor {
     ) -> Result<()> {
         let rtxn = &context.rtxn;
         let index = context.index;
-        let max_memory = self.grenad_parameters.max_memory;
+        let max_memory = self.grenad_parameters.max_memory_by_thread();
         let db_fields_ids_map = context.db_fields_ids_map;
         let mut data_ref = context.data.borrow_mut_or_yield();
 

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

@@ -13,9 +13,8 @@ pub use geo::*;
 pub use searchable::*;
 pub use vectors::EmbeddingExtractor;
 
-use super::indexer::document_changes::{
-    DocumentChanges, FullySend, IndexingContext, Progress, ThreadLocal,
-};
+use super::indexer::document_changes::{DocumentChanges, IndexingContext, Progress};
+use super::thread_local::{FullySend, ThreadLocal};
 use crate::update::GrenadParameters;
 use crate::Result;
 

crates/milli/src/update/new/extract/searchable/extract_word_docids.rs

@@ -11,10 +11,10 @@ use super::tokenize_document::{tokenizer_builder, DocumentTokenizer};
 use crate::update::new::extract::cache::BalancedCaches;
 use crate::update::new::extract::perm_json_p::contained_in;
 use crate::update::new::indexer::document_changes::{
-    extract, DocumentChangeContext, DocumentChanges, Extractor, FullySend, IndexingContext,
-    MostlySend, Progress, ThreadLocal,
+    extract, DocumentChangeContext, DocumentChanges, Extractor, IndexingContext, Progress,
 };
 use crate::update::new::ref_cell_ext::RefCellExt as _;
+use crate::update::new::thread_local::{FullySend, MostlySend, ThreadLocal};
 use crate::update::new::DocumentChange;
 use crate::update::GrenadParameters;
 use crate::{bucketed_position, DocumentId, FieldId, Index, Result, MAX_POSITION_PER_ATTRIBUTE};
@@ -214,7 +214,7 @@ impl<'a, 'extractor> Extractor<'extractor> for WordDocidsExtractorData<'a> {
     fn init_data(&self, extractor_alloc: &'extractor Bump) -> Result<Self::Data> {
         Ok(RefCell::new(Some(WordDocidsBalancedCaches::new_in(
             self.buckets,
-            self.grenad_parameters.max_memory,
+            self.grenad_parameters.max_memory_by_thread(),
             extractor_alloc,
         ))))
     }

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

@@ -14,9 +14,9 @@ use tokenize_document::{tokenizer_builder, DocumentTokenizer};
 use super::cache::BalancedCaches;
 use super::DocidsExtractor;
 use crate::update::new::indexer::document_changes::{
-    extract, DocumentChangeContext, DocumentChanges, Extractor, FullySend, IndexingContext,
-    Progress, ThreadLocal,
+    extract, DocumentChangeContext, DocumentChanges, Extractor, IndexingContext, Progress,
 };
+use crate::update::new::thread_local::{FullySend, ThreadLocal};
 use crate::update::new::DocumentChange;
 use crate::update::GrenadParameters;
 use crate::{Index, Result, MAX_POSITION_PER_ATTRIBUTE};
@@ -36,7 +36,7 @@ impl<'a, 'extractor, EX: SearchableExtractor + Sync> Extractor<'extractor>
     fn init_data(&self, extractor_alloc: &'extractor Bump) -> Result<Self::Data> {
         Ok(RefCell::new(BalancedCaches::new_in(
             self.buckets,
-            self.grenad_parameters.max_memory,
+            self.grenad_parameters.max_memory_by_thread(),
             extractor_alloc,
         )))
     }

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

@@ -8,7 +8,8 @@ use super::cache::DelAddRoaringBitmap;
 use crate::error::FaultSource;
 use crate::prompt::Prompt;
 use crate::update::new::channel::EmbeddingSender;
-use crate::update::new::indexer::document_changes::{DocumentChangeContext, Extractor, MostlySend};
+use crate::update::new::indexer::document_changes::{DocumentChangeContext, Extractor};
+use crate::update::new::thread_local::MostlySend;
 use crate::update::new::vector_document::VectorDocument;
 use crate::update::new::DocumentChange;
 use crate::vector::error::{

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

@@ -8,182 +8,9 @@ use rayon::iter::IndexedParallelIterator;
 use super::super::document_change::DocumentChange;
 use crate::fields_ids_map::metadata::FieldIdMapWithMetadata;
 use crate::update::new::parallel_iterator_ext::ParallelIteratorExt as _;
+use crate::update::new::thread_local::{FullySend, MostlySend, ThreadLocal};
 use crate::{FieldsIdsMap, GlobalFieldsIdsMap, Index, InternalError, Result};
 
-/// A trait for types that are **not** [`Send`] only because they would then allow concurrent access to a type that is not [`Sync`].
-///
-/// The primary example of such a type is `&T`, with `T: !Sync`.
-///
-/// In the authors' understanding, a type can be `!Send` for two distinct reasons:
-///
-/// 1. Because it contains data that *genuinely* cannot be moved between threads, such as thread-local data.
-/// 2. Because sending the type would allow concurrent access to a `!Sync` type, which is undefined behavior.
-///
-/// `MostlySend` exists to be used in bounds where you need a type whose data is **not** *attached* to a thread
-/// because you might access it from a different thread, but where you will never access the type **concurrently** from
-/// multiple threads.
-///
-/// Like [`Send`], `MostlySend` assumes properties on types that cannot be verified by the compiler, which is why implementing
-/// this trait is unsafe.
-///
-/// # Safety
-///
-/// Implementers of this trait promises that the following properties hold on the implementing type:
-///
-/// 1. Its data can be accessed from any thread and will be the same regardless of the thread accessing it.
-/// 2. Any operation that can be performed on the type does not depend on the thread that executes it.
-///
-/// As these properties are subtle and are not generally tracked by the Rust type system, great care should be taken before
-/// implementing `MostlySend` on a type, especially a foreign type.
-///
-/// - An example of a type that verifies (1) and (2) is [`std::rc::Rc`] (when `T` is `Send` and `Sync`).
-/// - An example of a type that doesn't verify (1) is thread-local data.
-/// - An example of a type that doesn't verify (2) is [`std::sync::MutexGuard`]: a lot of mutex implementations require that
-/// a lock is returned to the operating system on the same thread that initially locked the mutex, failing to uphold this
-/// invariant will cause Undefined Behavior
-/// (see last § in [the nomicon](https://doc.rust-lang.org/nomicon/send-and-sync.html)).
-///
-/// It is **always safe** to implement this trait on a type that is `Send`, but no placeholder impl is provided due to limitations in
-/// coherency. Use the [`FullySend`] wrapper in this situation.
-pub unsafe trait MostlySend {}
-
-#[derive(Debug, Clone, Copy, Default, Eq, PartialEq, Ord, PartialOrd, Hash)]
-pub struct FullySend<T>(pub T);
-
-// SAFETY: a type **fully** send is always mostly send as well.
-unsafe impl<T> MostlySend for FullySend<T> where T: Send {}
-
-unsafe impl<T> MostlySend for RefCell<T> where T: MostlySend {}
-
-unsafe impl<T> MostlySend for Option<T> where T: MostlySend {}
-
-impl<T> FullySend<T> {
-    pub fn into(self) -> T {
-        self.0
-    }
-}
-
-impl<T> From<T> for FullySend<T> {
-    fn from(value: T) -> Self {
-        Self(value)
-    }
-}
-
-#[derive(Debug, Clone, Copy, Default, PartialEq, Eq, PartialOrd, Ord, Hash)]
-struct MostlySendWrapper<T>(T);
-
-impl<T: MostlySend> MostlySendWrapper<T> {
-    /// # Safety
-    ///
-    /// - (P1) Users of this type will never access the type concurrently from multiple threads without synchronization
-    unsafe fn new(t: T) -> Self {
-        Self(t)
-    }
-
-    fn as_ref(&self) -> &T {
-        &self.0
-    }
-
-    fn as_mut(&mut self) -> &mut T {
-        &mut self.0
-    }
-
-    fn into_inner(self) -> T {
-        self.0
-    }
-}
-
-/// # Safety
-///
-/// 1. `T` is [`MostlySend`], so by its safety contract it can be accessed by any thread and all of its operations are available
-///   from any thread.
-/// 2. (P1) of `MostlySendWrapper::new` forces the user to never access the value from multiple threads concurrently.
-unsafe impl<T: MostlySend> Send for MostlySendWrapper<T> {}
-
-/// A wrapper around [`thread_local::ThreadLocal`] that accepts [`MostlySend`] `T`s.
-#[derive(Default)]
-pub struct ThreadLocal<T: MostlySend> {
-    inner: thread_local::ThreadLocal<MostlySendWrapper<T>>,
-    // FIXME: this should be necessary
-    //_no_send: PhantomData<*mut ()>,
-}
-
-impl<T: MostlySend> ThreadLocal<T> {
-    pub fn new() -> Self {
-        Self { inner: thread_local::ThreadLocal::new() }
-    }
-
-    pub fn with_capacity(capacity: usize) -> Self {
-        Self { inner: thread_local::ThreadLocal::with_capacity(capacity) }
-    }
-
-    pub fn clear(&mut self) {
-        self.inner.clear()
-    }
-
-    pub fn get(&self) -> Option<&T> {
-        self.inner.get().map(|t| t.as_ref())
-    }
-
-    pub fn get_or<F>(&self, create: F) -> &T
-    where
-        F: FnOnce() -> T,
-    {
-        /// TODO: move ThreadLocal, MostlySend, FullySend to a dedicated file
-        self.inner.get_or(|| unsafe { MostlySendWrapper::new(create()) }).as_ref()
-    }
-
-    pub fn get_or_try<F, E>(&self, create: F) -> std::result::Result<&T, E>
-    where
-        F: FnOnce() -> std::result::Result<T, E>,
-    {
-        self.inner
-            .get_or_try(|| unsafe { Ok(MostlySendWrapper::new(create()?)) })
-            .map(MostlySendWrapper::as_ref)
-    }
-
-    pub fn get_or_default(&self) -> &T
-    where
-        T: Default,
-    {
-        self.inner.get_or_default().as_ref()
-    }
-
-    pub fn iter_mut(&mut self) -> IterMut<T> {
-        IterMut(self.inner.iter_mut())
-    }
-}
-
-impl<T: MostlySend> IntoIterator for ThreadLocal<T> {
-    type Item = T;
-
-    type IntoIter = IntoIter<T>;
-
-    fn into_iter(self) -> Self::IntoIter {
-        IntoIter(self.inner.into_iter())
-    }
-}
-
-pub struct IterMut<'a, T: MostlySend>(thread_local::IterMut<'a, MostlySendWrapper<T>>);
-
-impl<'a, T: MostlySend> Iterator for IterMut<'a, T> {
-    type Item = &'a mut T;
-
-    fn next(&mut self) -> Option<Self::Item> {
-        self.0.next().map(|t| t.as_mut())
-    }
-}
-
-pub struct IntoIter<T: MostlySend>(thread_local::IntoIter<MostlySendWrapper<T>>);
-
-impl<T: MostlySend> Iterator for IntoIter<T> {
-    type Item = T;
-
-    fn next(&mut self) -> Option<Self::Item> {
-        self.0.next().map(|t| t.into_inner())
-    }
-}
-
 pub struct DocumentChangeContext<
     'doc,             // covariant lifetime of a single `process` call
     'extractor: 'doc, // invariant lifetime of the extractor_allocs

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

@@ -4,8 +4,9 @@ use rayon::iter::IndexedParallelIterator;
 use rayon::slice::ParallelSlice as _;
 use roaring::RoaringBitmap;
 
-use super::document_changes::{DocumentChangeContext, DocumentChanges, MostlySend};
+use super::document_changes::{DocumentChangeContext, DocumentChanges};
 use crate::documents::PrimaryKey;
+use crate::update::new::thread_local::MostlySend;
 use crate::update::new::{Deletion, DocumentChange};
 use crate::{DocumentId, Result};
 
@@ -92,9 +93,10 @@ mod test {
     use crate::fields_ids_map::metadata::{FieldIdMapWithMetadata, MetadataBuilder};
     use crate::index::tests::TempIndex;
     use crate::update::new::indexer::document_changes::{
-        extract, DocumentChangeContext, Extractor, IndexingContext, MostlySend, ThreadLocal,
+        extract, DocumentChangeContext, Extractor, IndexingContext,
     };
     use crate::update::new::indexer::DocumentDeletion;
+    use crate::update::new::thread_local::{MostlySend, ThreadLocal};
     use crate::update::new::DocumentChange;
     use crate::DocumentId;
 

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

@@ -1,66 +1,39 @@
 use bumpalo::collections::CollectIn;
 use bumpalo::Bump;
+use hashbrown::hash_map::Entry;
 use heed::RoTxn;
 use memmap2::Mmap;
+use raw_collections::RawMap;
 use rayon::slice::ParallelSlice;
 use serde_json::value::RawValue;
-use IndexDocumentsMethod as Idm;
+use serde_json::Deserializer;
 
 use super::super::document_change::DocumentChange;
-use super::document_changes::{DocumentChangeContext, DocumentChanges, MostlySend};
+use super::document_changes::{DocumentChangeContext, DocumentChanges};
+use super::retrieve_or_guess_primary_key;
 use crate::documents::PrimaryKey;
 use crate::update::new::document::Versions;
+use crate::update::new::thread_local::MostlySend;
 use crate::update::new::{Deletion, Insertion, Update};
 use crate::update::{AvailableIds, IndexDocumentsMethod};
-use crate::{DocumentId, Error, FieldsIdsMap, Index, Result, UserError};
+use crate::{DocumentId, Error, FieldsIdsMap, Index, InternalError, Result, UserError};
 
 pub struct DocumentOperation<'pl> {
     operations: Vec<Payload<'pl>>,
-    index_documents_method: IndexDocumentsMethod,
-}
-
-pub struct DocumentOperationChanges<'pl> {
-    docids_version_offsets: &'pl [(&'pl str, ((u32, bool), &'pl [InnerDocOp<'pl>]))],
-    index_documents_method: IndexDocumentsMethod,
-}
-
-pub enum Payload<'pl> {
-    Addition(&'pl [u8]),
-    Deletion(&'pl [&'pl str]),
-}
-
-pub struct PayloadStats {
-    pub document_count: usize,
-    pub bytes: u64,
-}
-
-#[derive(Clone)]
-pub enum InnerDocOp<'pl> {
-    Addition(DocumentOffset<'pl>),
-    Deletion,
-}
-
-/// Represents an offset where a document lives
-/// in an mmapped grenad reader file.
-#[derive(Clone)]
-pub struct DocumentOffset<'pl> {
-    /// The mmapped payload files.
-    pub content: &'pl [u8],
+    method: MergeMethod,
 }
 
 impl<'pl> DocumentOperation<'pl> {
     pub fn new(method: IndexDocumentsMethod) -> Self {
-        Self { operations: Default::default(), index_documents_method: method }
+        Self { operations: Default::default(), method: MergeMethod::from(method) }
     }
 
     /// TODO please give me a type
     /// The payload is expected to be in the grenad format
-    pub fn add_documents(&mut self, payload: &'pl Mmap) -> Result<PayloadStats> {
+    pub fn add_documents(&mut self, payload: &'pl Mmap) -> Result<()> {
         payload.advise(memmap2::Advice::Sequential)?;
-        let document_count =
-            memchr::memmem::find_iter(&payload[..], "}{").count().saturating_add(1);
         self.operations.push(Payload::Addition(&payload[..]));
-        Ok(PayloadStats { bytes: payload.len() as u64, document_count })
+        Ok(())
     }
 
     pub fn delete_documents(&mut self, to_delete: &'pl [&'pl str]) {
@@ -71,144 +44,301 @@ impl<'pl> DocumentOperation<'pl> {
         self,
         indexer: &'pl Bump,
         index: &Index,
-        rtxn: &RoTxn,
-        primary_key: &PrimaryKey,
+        rtxn: &'pl RoTxn<'pl>,
+        primary_key_from_op: Option<&'pl str>,
         new_fields_ids_map: &mut FieldsIdsMap,
-    ) -> Result<DocumentOperationChanges<'pl>> {
-        // will contain nodes from the intermediate hashmap
-        let document_changes_alloc = Bump::with_capacity(1024 * 1024 * 1024); // 1 MiB
+    ) -> Result<(DocumentOperationChanges<'pl>, Vec<PayloadStats>, Option<PrimaryKey<'pl>>)> {
+        let Self { operations, method } = self;
 
         let documents_ids = index.documents_ids(rtxn)?;
+        let mut operations_stats = Vec::new();
         let mut available_docids = AvailableIds::new(&documents_ids);
-        let mut docids_version_offsets =
-            hashbrown::HashMap::<&'pl str, _, _, _>::new_in(&document_changes_alloc);
+        let mut docids_version_offsets = hashbrown::HashMap::new();
+        let mut primary_key = None;
 
-        for operation in self.operations {
-            match operation {
-                Payload::Addition(payload) => {
-                    let mut iter =
-                        serde_json::Deserializer::from_slice(payload).into_iter::<&RawValue>();
+        for operation in operations {
+            let (bytes, document_count, result) = match operation {
+                Payload::Addition(payload) => extract_addition_payload_changes(
+                    indexer,
+                    index,
+                    rtxn,
+                    primary_key_from_op,
+                    &mut primary_key,
+                    new_fields_ids_map,
+                    &mut available_docids,
+                    &docids_version_offsets,
+                    method,
+                    payload,
+                ),
+                Payload::Deletion(to_delete) => extract_deletion_payload_changes(
+                    index,
+                    rtxn,
+                    &mut available_docids,
+                    &docids_version_offsets,
+                    method,
+                    to_delete,
+                ),
+            };
+
+            let error = match result {
+                Ok(new_docids_version_offsets) => {
+                    // If we don't have any error then we can merge the content of this payload
+                    // into to main payload. Else we just drop this payload extraction.
+                    merge_version_offsets(&mut docids_version_offsets, new_docids_version_offsets);
+                    None
+                }
+                Err(Error::UserError(user_error)) => Some(user_error),
+                Err(e) => return Err(e),
+            };
+
+            operations_stats.push(PayloadStats { document_count, bytes, error });
+        }
+
+        // TODO We must drain the HashMap into a Vec because rayon::hash_map::IntoIter: !Clone
+        let mut docids_version_offsets: bumpalo::collections::vec::Vec<_> =
+            docids_version_offsets.drain().collect_in(indexer);
+
+        // Reorder the offsets to make sure we iterate on the file sequentially
+        // And finally sort them
+        docids_version_offsets.sort_unstable_by_key(|(_, po)| method.sort_key(&po.operations));
+
+        let docids_version_offsets = docids_version_offsets.into_bump_slice();
+        Ok((DocumentOperationChanges { docids_version_offsets }, operations_stats, primary_key))
+    }
+}
+
+#[allow(clippy::too_many_arguments)]
+fn extract_addition_payload_changes<'r, 'pl: 'r>(
+    indexer: &'pl Bump,
+    index: &Index,
+    rtxn: &'r RoTxn<'r>,
+    primary_key_from_op: Option<&'r str>,
+    primary_key: &mut Option<PrimaryKey<'r>>,
+    new_fields_ids_map: &mut FieldsIdsMap,
+    available_docids: &mut AvailableIds,
+    main_docids_version_offsets: &hashbrown::HashMap<&'pl str, PayloadOperations<'pl>>,
+    method: MergeMethod,
+    payload: &'pl [u8],
+) -> (u64, u64, Result<hashbrown::HashMap<&'pl str, PayloadOperations<'pl>>>) {
+    let mut new_docids_version_offsets = hashbrown::HashMap::<&str, PayloadOperations<'pl>>::new();
 
     /// TODO manage the error
     let mut previous_offset = 0;
-                    while let Some(document) =
-                        iter.next().transpose().map_err(UserError::SerdeJson)?
-                    {
-                        let external_document_id = primary_key.extract_fields_and_docid(
-                            document,
+    let mut iter = Deserializer::from_slice(payload).into_iter::<&RawValue>();
+    loop {
+        let optdoc = match iter.next().transpose() {
+            Ok(optdoc) => optdoc,
+            Err(e) => {
+                return (
+                    payload.len() as u64,
+                    new_docids_version_offsets.len() as u64,
+                    Err(InternalError::SerdeJson(e).into()),
+                )
+            }
+        };
+
+        // Only guess the primary key if it is the first document
+        let retrieved_primary_key = if previous_offset == 0 {
+            let optdoc = match optdoc {
+                Some(doc) => match RawMap::from_raw_value(doc, indexer) {
+                    Ok(docmap) => Some(docmap),
+                    Err(error) => {
+                        return (
+                            payload.len() as u64,
+                            new_docids_version_offsets.len() as u64,
+                            Err(Error::UserError(UserError::SerdeJson(error))),
+                        )
+                    }
+                },
+                None => None,
+            };
+
+            let result = retrieve_or_guess_primary_key(
+                rtxn,
+                index,
+                new_fields_ids_map,
+                primary_key_from_op,
+                optdoc,
+            );
+
+            let (pk, _has_been_changed) = match result {
+                Ok(Ok(pk)) => pk,
+                Ok(Err(user_error)) => {
+                    return (
+                        payload.len() as u64,
+                        new_docids_version_offsets.len() as u64,
+                        Err(Error::UserError(user_error)),
+                    )
+                }
+                Err(error) => {
+                    return (
+                        payload.len() as u64,
+                        new_docids_version_offsets.len() as u64,
+                        Err(error),
+                    )
+                }
+            };
+
+            primary_key.get_or_insert(pk)
+        } else {
+            primary_key.as_ref().unwrap()
+        };
+
+        let doc = match optdoc {
+            Some(doc) => doc,
+            None => break,
+        };
+
+        let external_id = match retrieved_primary_key.extract_fields_and_docid(
+            doc,
             new_fields_ids_map,
             indexer,
-                        )?;
-
-                        let external_document_id = external_document_id.to_de();
+        ) {
+            Ok(edi) => edi,
+            Err(e) => {
+                return (payload.len() as u64, new_docids_version_offsets.len() as u64, Err(e))
+            }
+        };
 
+        let external_id = external_id.to_de();
         let current_offset = iter.byte_offset();
-                        let document_operation = InnerDocOp::Addition(DocumentOffset {
-                            content: &payload[previous_offset..current_offset],
-                        });
+        let document_offset = DocumentOffset { content: &payload[previous_offset..current_offset] };
 
-                        match docids_version_offsets.get_mut(external_document_id) {
+        match main_docids_version_offsets.get(external_id) {
             None => {
-                                let (docid, is_new) = match index
-                                    .external_documents_ids()
-                                    .get(rtxn, external_document_id)?
-                                {
-                                    Some(docid) => (docid, false),
-                                    None => (
-                                        available_docids.next().ok_or(Error::UserError(
-                                            UserError::DocumentLimitReached,
-                                        ))?,
+                let (docid, is_new) = match index.external_documents_ids().get(rtxn, external_id) {
+                    Ok(Some(docid)) => (docid, false),
+                    Ok(None) => (
+                        match available_docids.next() {
+                            Some(docid) => docid,
+                            None => {
+                                return (
+                                    payload.len() as u64,
+                                    new_docids_version_offsets.len() as u64,
+                                    Err(UserError::DocumentLimitReached.into()),
+                                )
+                            }
+                        },
                         true,
                     ),
-                                };
-
-                                docids_version_offsets.insert(
-                                    external_document_id,
-                                    (
-                                        (docid, is_new),
-                                        bumpalo::vec![in indexer; document_operation],
-                                    ),
-                                );
-                            }
-                            Some((_, offsets)) => {
-                                let useless_previous_addition = match self.index_documents_method {
-                                    IndexDocumentsMethod::ReplaceDocuments => {
-                                        MergeDocumentForReplacement::USELESS_PREVIOUS_CHANGES
-                                    }
-                                    IndexDocumentsMethod::UpdateDocuments => {
-                                        MergeDocumentForUpdates::USELESS_PREVIOUS_CHANGES
+                    Err(e) => {
+                        return (
+                            payload.len() as u64,
+                            new_docids_version_offsets.len() as u64,
+                            Err(e.into()),
+                        )
                     }
                 };
 
-                                if useless_previous_addition {
-                                    offsets.clear();
+                match new_docids_version_offsets.entry(external_id) {
+                    Entry::Occupied(mut entry) => entry.get_mut().push_addition(document_offset),
+                    Entry::Vacant(entry) => {
+                        entry.insert(PayloadOperations::new_addition(
+                            method,
+                            docid,
+                            is_new,
+                            document_offset,
+                        ));
                     }
-
-                                offsets.push(document_operation);
                 }
             }
+            Some(payload_operations) => match new_docids_version_offsets.entry(external_id) {
+                Entry::Occupied(mut entry) => entry.get_mut().push_addition(document_offset),
+                Entry::Vacant(entry) => {
+                    entry.insert(PayloadOperations::new_addition(
+                        method,
+                        payload_operations.docid,
+                        payload_operations.is_new,
+                        document_offset,
+                    ));
+                }
+            },
+        }
 
         previous_offset = iter.byte_offset();
     }
-                }
-                Payload::Deletion(to_delete) => {
-                    for external_document_id in to_delete {
-                        match docids_version_offsets.get_mut(external_document_id) {
+
+    (payload.len() as u64, new_docids_version_offsets.len() as u64, Ok(new_docids_version_offsets))
+}
+
+fn extract_deletion_payload_changes<'s, 'pl: 's>(
+    index: &Index,
+    rtxn: &RoTxn,
+    available_docids: &mut AvailableIds,
+    main_docids_version_offsets: &hashbrown::HashMap<&'s str, PayloadOperations<'pl>>,
+    method: MergeMethod,
+    to_delete: &'pl [&'pl str],
+) -> (u64, u64, Result<hashbrown::HashMap<&'s str, PayloadOperations<'pl>>>) {
+    let mut new_docids_version_offsets = hashbrown::HashMap::<&str, PayloadOperations<'pl>>::new();
+    let mut document_count = 0;
+
+    for external_id in to_delete {
+        match main_docids_version_offsets.get(external_id) {
             None => {
-                                let (docid, is_new) = match index
-                                    .external_documents_ids()
-                                    .get(rtxn, external_document_id)?
-                                {
-                                    Some(docid) => (docid, false),
-                                    None => (
-                                        available_docids.next().ok_or(Error::UserError(
-                                            UserError::DocumentLimitReached,
-                                        ))?,
+                let (docid, is_new) = match index.external_documents_ids().get(rtxn, external_id) {
+                    Ok(Some(docid)) => (docid, false),
+                    Ok(None) => (
+                        match available_docids.next() {
+                            Some(docid) => docid,
+                            None => {
+                                return (
+                                    0,
+                                    new_docids_version_offsets.len() as u64,
+                                    Err(UserError::DocumentLimitReached.into()),
+                                )
+                            }
+                        },
                         true,
                     ),
+                    Err(e) => return (0, new_docids_version_offsets.len() as u64, Err(e.into())),
                 };
 
-                                docids_version_offsets.insert(
-                                    external_document_id,
-                                    (
-                                        (docid, is_new),
-                                        bumpalo::vec![in indexer; InnerDocOp::Deletion],
-                                    ),
-                                );
-                            }
-                            Some((_, offsets)) => {
-                                offsets.clear();
-                                offsets.push(InnerDocOp::Deletion);
+                match new_docids_version_offsets.entry(external_id) {
+                    Entry::Occupied(mut entry) => entry.get_mut().push_deletion(),
+                    Entry::Vacant(entry) => {
+                        entry.insert(PayloadOperations::new_deletion(method, docid, is_new));
                     }
                 }
             }
+            Some(payload_operations) => match new_docids_version_offsets.entry(external_id) {
+                Entry::Occupied(mut entry) => entry.get_mut().push_deletion(),
+                Entry::Vacant(entry) => {
+                    entry.insert(PayloadOperations::new_deletion(
+                        method,
+                        payload_operations.docid,
+                        payload_operations.is_new,
+                    ));
                 }
+            },
         }
+        document_count += 1;
     }
 
-        // TODO We must drain the HashMap into a Vec because rayon::hash_map::IntoIter: !Clone
-        let mut docids_version_offsets: bumpalo::collections::vec::Vec<_> = docids_version_offsets
-            .drain()
-            .map(|(item, (docid, v))| (item, (docid, v.into_bump_slice())))
-            .collect_in(indexer);
-        // Reorder the offsets to make sure we iterate on the file sequentially
-        let sort_function_key = match self.index_documents_method {
-            Idm::ReplaceDocuments => MergeDocumentForReplacement::sort_key,
-            Idm::UpdateDocuments => MergeDocumentForUpdates::sort_key,
-        };
+    (0, document_count, Ok(new_docids_version_offsets))
+}
 
-        // And finally sort them
-        docids_version_offsets.sort_unstable_by_key(|(_, (_, docops))| sort_function_key(docops));
-        let docids_version_offsets = docids_version_offsets.into_bump_slice();
-        Ok(DocumentOperationChanges {
-            docids_version_offsets,
-            index_documents_method: self.index_documents_method,
-        })
+fn merge_version_offsets<'s, 'pl>(
+    main: &mut hashbrown::HashMap<&'s str, PayloadOperations<'pl>>,
+    new: hashbrown::HashMap<&'s str, PayloadOperations<'pl>>,
+) {
+    // We cannot swap like nothing because documents
+    // operations must be in the right order.
+    if main.is_empty() {
+        return *main = new;
+    }
+
+    for (key, new_payload) in new {
+        match main.entry(key) {
+            Entry::Occupied(mut entry) => entry.get_mut().append_operations(new_payload.operations),
+            Entry::Vacant(entry) => {
+                entry.insert(new_payload);
+            }
+        }
     }
 }
 
 impl<'pl> DocumentChanges<'pl> for DocumentOperationChanges<'pl> {
-    type Item = (&'pl str, ((u32, bool), &'pl [InnerDocOp<'pl>]));
+    type Item = (&'pl str, PayloadOperations<'pl>);
 
     fn iter(
         &self,
@@ -225,21 +355,14 @@ impl<'pl> DocumentChanges<'pl> for DocumentOperationChanges<'pl> {
     where
         'pl: 'doc,
     {
-        let document_merge_function = match self.index_documents_method {
-            Idm::ReplaceDocuments => MergeDocumentForReplacement::merge,
-            Idm::UpdateDocuments => MergeDocumentForUpdates::merge,
-        };
-
-        let (external_doc, ((internal_docid, is_new), operations)) = *item;
-
-        let change = document_merge_function(
-            internal_docid,
+        let (external_doc, payload_operations) = item;
+        payload_operations.merge_method.merge(
+            payload_operations.docid,
             external_doc,
-            is_new,
+            payload_operations.is_new,
             &context.doc_alloc,
-            operations,
-        )?;
-        Ok(change)
+            &payload_operations.operations[..],
+        )
     }
 
     fn len(&self) -> usize {
@@ -247,14 +370,92 @@ impl<'pl> DocumentChanges<'pl> for DocumentOperationChanges<'pl> {
     }
 }
 
+pub struct DocumentOperationChanges<'pl> {
+    docids_version_offsets: &'pl [(&'pl str, PayloadOperations<'pl>)],
+}
+
+pub enum Payload<'pl> {
+    Addition(&'pl [u8]),
+    Deletion(&'pl [&'pl str]),
+}
+
+pub struct PayloadStats {
+    pub bytes: u64,
+    pub document_count: u64,
+    pub error: Option<UserError>,
+}
+
+pub struct PayloadOperations<'pl> {
+    /// The internal document id of the document.
+    pub docid: DocumentId,
+    /// Wether this document is not in the current database (visible by the rtxn).
+    pub is_new: bool,
+    /// The operations to perform, in order, on this document.
+    pub operations: Vec<InnerDocOp<'pl>>,
+    /// The merge method we are using to merge payloads and documents.
+    merge_method: MergeMethod,
+}
+
+impl<'pl> PayloadOperations<'pl> {
+    fn new_deletion(merge_method: MergeMethod, docid: DocumentId, is_new: bool) -> Self {
+        Self { docid, is_new, operations: vec![InnerDocOp::Deletion], merge_method }
+    }
+
+    fn new_addition(
+        merge_method: MergeMethod,
+        docid: DocumentId,
+        is_new: bool,
+        offset: DocumentOffset<'pl>,
+    ) -> Self {
+        Self { docid, is_new, operations: vec![InnerDocOp::Addition(offset)], merge_method }
+    }
+}
+
+impl<'pl> PayloadOperations<'pl> {
+    fn push_addition(&mut self, offset: DocumentOffset<'pl>) {
+        if self.merge_method.useless_previous_changes() {
+            self.operations.clear();
+        }
+        self.operations.push(InnerDocOp::Addition(offset))
+    }
+
+    fn push_deletion(&mut self) {
+        self.operations.clear();
+        self.operations.push(InnerDocOp::Deletion);
+    }
+
+    fn append_operations(&mut self, mut operations: Vec<InnerDocOp<'pl>>) {
+        debug_assert!(!operations.is_empty());
+        if self.merge_method.useless_previous_changes() {
+            self.operations.clear();
+        }
+        self.operations.append(&mut operations);
+    }
+}
+
+#[derive(Clone)]
+pub enum InnerDocOp<'pl> {
+    Addition(DocumentOffset<'pl>),
+    Deletion,
+}
+
+/// Represents an offset where a document lives
+/// in an mmapped grenad reader file.
+#[derive(Clone)]
+pub struct DocumentOffset<'pl> {
+    /// The mmapped payload files.
+    pub content: &'pl [u8],
+}
+
 trait MergeChanges {
     /// Whether the payloads in the list of operations are useless or not.
-    const USELESS_PREVIOUS_CHANGES: bool;
+    fn useless_previous_changes(&self) -> bool;
 
     /// Returns a key that is used to order the payloads the right way.
-    fn sort_key(docops: &[InnerDocOp]) -> usize;
+    fn sort_key(&self, docops: &[InnerDocOp]) -> usize;
 
     fn merge<'doc>(
+        &self,
         docid: DocumentId,
         external_docid: &'doc str,
         is_new: bool,
@@ -263,13 +464,69 @@ trait MergeChanges {
     ) -> Result<Option<DocumentChange<'doc>>>;
 }
 
+#[derive(Debug, Clone, Copy)]
+enum MergeMethod {
+    ForReplacement(MergeDocumentForReplacement),
+    ForUpdates(MergeDocumentForUpdates),
+}
+
+impl MergeChanges for MergeMethod {
+    fn useless_previous_changes(&self) -> bool {
+        match self {
+            MergeMethod::ForReplacement(merge) => merge.useless_previous_changes(),
+            MergeMethod::ForUpdates(merge) => merge.useless_previous_changes(),
+        }
+    }
+
+    fn sort_key(&self, docops: &[InnerDocOp]) -> usize {
+        match self {
+            MergeMethod::ForReplacement(merge) => merge.sort_key(docops),
+            MergeMethod::ForUpdates(merge) => merge.sort_key(docops),
+        }
+    }
+
+    fn merge<'doc>(
+        &self,
+        docid: DocumentId,
+        external_docid: &'doc str,
+        is_new: bool,
+        doc_alloc: &'doc Bump,
+        operations: &'doc [InnerDocOp],
+    ) -> Result<Option<DocumentChange<'doc>>> {
+        match self {
+            MergeMethod::ForReplacement(merge) => {
+                merge.merge(docid, external_docid, is_new, doc_alloc, operations)
+            }
+            MergeMethod::ForUpdates(merge) => {
+                merge.merge(docid, external_docid, is_new, doc_alloc, operations)
+            }
+        }
+    }
+}
+
+impl From<IndexDocumentsMethod> for MergeMethod {
+    fn from(method: IndexDocumentsMethod) -> Self {
+        match method {
+            IndexDocumentsMethod::ReplaceDocuments => {
+                MergeMethod::ForReplacement(MergeDocumentForReplacement)
+            }
+            IndexDocumentsMethod::UpdateDocuments => {
+                MergeMethod::ForUpdates(MergeDocumentForUpdates)
+            }
+        }
+    }
+}
+
+#[derive(Debug, Clone, Copy)]
 struct MergeDocumentForReplacement;
 
 impl MergeChanges for MergeDocumentForReplacement {
-    const USELESS_PREVIOUS_CHANGES: bool = true;
+    fn useless_previous_changes(&self) -> bool {
+        true
+    }
 
     /// Reorders to read only the last change.
-    fn sort_key(docops: &[InnerDocOp]) -> usize {
+    fn sort_key(&self, docops: &[InnerDocOp]) -> usize {
         let f = |ido: &_| match ido {
             InnerDocOp::Addition(add) => Some(add.content.as_ptr() as usize),
             InnerDocOp::Deletion => None,
@@ -281,6 +538,7 @@ impl MergeChanges for MergeDocumentForReplacement {
     ///
     /// This function is only meant to be used when doing a replacement and not an update.
     fn merge<'doc>(
+        &self,
         docid: DocumentId,
         external_doc: &'doc str,
         is_new: bool,
@@ -321,13 +579,16 @@ impl MergeChanges for MergeDocumentForReplacement {
     }
 }
 
+#[derive(Debug, Clone, Copy)]
 struct MergeDocumentForUpdates;
 
 impl MergeChanges for MergeDocumentForUpdates {
-    const USELESS_PREVIOUS_CHANGES: bool = false;
+    fn useless_previous_changes(&self) -> bool {
+        false
+    }
 
     /// Reorders to read the first changes first so that it's faster to read the first one and then the rest.
-    fn sort_key(docops: &[InnerDocOp]) -> usize {
+    fn sort_key(&self, docops: &[InnerDocOp]) -> usize {
         let f = |ido: &_| match ido {
             InnerDocOp::Addition(add) => Some(add.content.as_ptr() as usize),
             InnerDocOp::Deletion => None,
@@ -340,6 +601,7 @@ impl MergeChanges for MergeDocumentForUpdates {
     ///
     /// This function is only meant to be used when doing an update and not a replacement.
     fn merge<'doc>(
+        &self,
         docid: DocumentId,
         external_docid: &'doc str,
         is_new: bool,

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

@@ -3,9 +3,9 @@ use std::sync::{OnceLock, RwLock};
 use std::thread::{self, Builder};
 
 use big_s::S;
-use document_changes::{extract, DocumentChanges, IndexingContext, Progress, ThreadLocal};
+use document_changes::{extract, DocumentChanges, IndexingContext, Progress};
 pub use document_deletion::DocumentDeletion;
-pub use document_operation::DocumentOperation;
+pub use document_operation::{DocumentOperation, PayloadStats};
 use hashbrown::HashMap;
 use heed::types::{Bytes, DecodeIgnore, Str};
 use heed::{RoTxn, RwTxn};
@@ -20,6 +20,7 @@ use super::channel::*;
 use super::extract::*;
 use super::facet_search_builder::FacetSearchBuilder;
 use super::merger::FacetFieldIdsDelta;
+use super::thread_local::ThreadLocal;
 use super::word_fst_builder::{PrefixData, PrefixDelta, WordFstBuilder};
 use super::words_prefix_docids::{
     compute_word_prefix_docids, compute_word_prefix_fid_docids, compute_word_prefix_position_docids,
@@ -132,6 +133,7 @@ mod steps {
 pub fn index<'pl, 'indexer, 'index, DC, MSP, SP>(
     wtxn: &mut RwTxn,
     index: &'index Index,
+    grenad_parameters: GrenadParameters,
     db_fields_ids_map: &'indexer FieldsIdsMap,
     new_fields_ids_map: FieldsIdsMap,
     new_primary_key: Option<PrimaryKey<'pl>>,
@@ -209,16 +211,6 @@ where
 
             field_distribution.retain(|_, v| *v != 0);
 
-            const TEN_GIB: usize = 10 * 1024 * 1024 * 1024;
-            let current_num_threads = rayon::current_num_threads();
-            let max_memory = TEN_GIB / current_num_threads;
-            eprintln!("A maximum of {max_memory} bytes will be used for each of the {current_num_threads} threads");
-
-            let grenad_parameters = GrenadParameters {
-                max_memory: Some(max_memory),
-                ..GrenadParameters::default()
-            };
-
             let facet_field_ids_delta;
 
             {
@@ -228,7 +220,8 @@ where
                 let (finished_steps, step_name) = steps::extract_facets();
 
                 facet_field_ids_delta = merge_and_send_facet_docids(
-                    FacetedDocidsExtractor::run_extraction(grenad_parameters,
+                    FacetedDocidsExtractor::run_extraction(
+                        grenad_parameters,
                         document_changes,
                         indexing_context,
                         &mut extractor_allocs,
@@ -344,7 +337,8 @@ where
 
                 let (finished_steps, step_name) = steps::extract_word_proximity();
 
-                let caches = <WordPairProximityDocidsExtractor as DocidsExtractor>::run_extraction(grenad_parameters,
+                let caches = <WordPairProximityDocidsExtractor as DocidsExtractor>::run_extraction(
+                    grenad_parameters,
                     document_changes,
                     indexing_context,
                     &mut extractor_allocs,
@@ -398,7 +392,8 @@ where
                 };
                 let datastore = ThreadLocal::with_capacity(rayon::current_num_threads());
                 let (finished_steps, step_name) = steps::extract_geo_points();
-                extract(document_changes,
+                extract(
+                    document_changes,
                     &extractor,
                     indexing_context,
                     &mut extractor_allocs,
@@ -777,16 +772,10 @@ pub fn retrieve_or_guess_primary_key<'a>(
         match primary_key_from_op {
             // we did, and it is different from the DB one
             Some(primary_key_from_op) if primary_key_from_op != primary_key_from_db => {
-                // is the index empty?
-                if index.number_of_documents(rtxn)? == 0 {
-                    // change primary key
-                    (primary_key_from_op, true)
-                } else {
                 return Ok(Err(UserError::PrimaryKeyCannotBeChanged(
                     primary_key_from_db.to_string(),
                 )));
             }
-            }
             _ => (primary_key_from_db, false),
         }
     } else {

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

@@ -3,11 +3,12 @@ use std::ops::DerefMut;
 use rayon::iter::IndexedParallelIterator;
 use serde_json::value::RawValue;
 
-use super::document_changes::{DocumentChangeContext, DocumentChanges, MostlySend};
+use super::document_changes::{DocumentChangeContext, DocumentChanges};
 use crate::documents::PrimaryKey;
 use crate::update::concurrent_available_ids::ConcurrentAvailableIds;
 use crate::update::new::document::Versions;
 use crate::update::new::ref_cell_ext::RefCellExt as _;
+use crate::update::new::thread_local::MostlySend;
 use crate::update::new::{DocumentChange, Insertion};
 use crate::{Error, InternalError, Result, UserError};
 

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

@@ -4,13 +4,14 @@ use rayon::slice::ParallelSlice as _;
 use rhai::{Dynamic, Engine, OptimizationLevel, Scope, AST};
 use roaring::RoaringBitmap;
 
-use super::document_changes::{DocumentChangeContext, MostlySend};
+use super::document_changes::DocumentChangeContext;
 use super::DocumentChanges;
 use crate::documents::Error::InvalidDocumentFormat;
 use crate::documents::PrimaryKey;
 use crate::error::{FieldIdMapMissingEntry, InternalError};
 use crate::update::new::document::Versions;
 use crate::update::new::ref_cell_ext::RefCellExt as _;
+use crate::update::new::thread_local::MostlySend;
 use crate::update::new::{Deletion, DocumentChange, KvReaderFieldId, Update};
 use crate::{all_obkv_to_json, Error, FieldsIdsMap, Object, Result, UserError};
 

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

@@ -17,6 +17,7 @@ pub mod indexer;
 mod merger;
 mod parallel_iterator_ext;
 mod ref_cell_ext;
+pub(crate) mod thread_local;
 mod top_level_map;
 pub mod vector_document;
 mod word_fst_builder;

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

@@ -1,37 +1,16 @@
-use std::cell::{Ref, RefCell, RefMut};
+use std::cell::{RefCell, RefMut};
 
 pub trait RefCellExt<T: ?Sized> {
-    fn try_borrow_or_yield(&self) -> std::result::Result<Ref<'_, T>, std::cell::BorrowError>;
     fn try_borrow_mut_or_yield(
         &self,
     ) -> std::result::Result<RefMut<'_, T>, std::cell::BorrowMutError>;
 
-    fn borrow_or_yield(&self) -> Ref<'_, T> {
-        self.try_borrow_or_yield().unwrap()
-    }
-
     fn borrow_mut_or_yield(&self) -> RefMut<'_, T> {
         self.try_borrow_mut_or_yield().unwrap()
     }
 }
 
 impl<T: ?Sized> RefCellExt<T> for RefCell<T> {
-    fn try_borrow_or_yield(&self) -> std::result::Result<Ref<'_, T>, std::cell::BorrowError> {
-        /// TODO: move this trait and impl elsewhere
-        loop {
-            match self.try_borrow() {
-                Ok(borrow) => break Ok(borrow),
-                Err(error) => {
-                    tracing::warn!("dynamic borrow failed, yielding to local tasks");
-                    match rayon::yield_local() {
-                        Some(rayon::Yield::Executed) => continue,
-                        _ => return Err(error),
-                    }
-                }
-            }
-        }
-    }
-
     fn try_borrow_mut_or_yield(
         &self,
     ) -> std::result::Result<RefMut<'_, T>, std::cell::BorrowMutError> {

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

@@ -0,0 +1,174 @@
+use std::cell::RefCell;
+
+/// A trait for types that are **not** [`Send`] only because they would then allow concurrent access to a type that is not [`Sync`].
+///
+/// The primary example of such a type is `&T`, with `T: !Sync`.
+///
+/// In the authors' understanding, a type can be `!Send` for two distinct reasons:
+///
+/// 1. Because it contains data that *genuinely* cannot be moved between threads, such as thread-local data.
+/// 2. Because sending the type would allow concurrent access to a `!Sync` type, which is undefined behavior.
+///
+/// `MostlySend` exists to be used in bounds where you need a type whose data is **not** *attached* to a thread
+/// because you might access it from a different thread, but where you will never access the type **concurrently** from
+/// multiple threads.
+///
+/// Like [`Send`], `MostlySend` assumes properties on types that cannot be verified by the compiler, which is why implementing
+/// this trait is unsafe.
+///
+/// # Safety
+///
+/// Implementers of this trait promises that the following properties hold on the implementing type:
+///
+/// 1. Its data can be accessed from any thread and will be the same regardless of the thread accessing it.
+/// 2. Any operation that can be performed on the type does not depend on the thread that executes it.
+///
+/// As these properties are subtle and are not generally tracked by the Rust type system, great care should be taken before
+/// implementing `MostlySend` on a type, especially a foreign type.
+///
+/// - An example of a type that verifies (1) and (2) is [`std::rc::Rc`] (when `T` is `Send` and `Sync`).
+/// - An example of a type that doesn't verify (1) is thread-local data.
+/// - An example of a type that doesn't verify (2) is [`std::sync::MutexGuard`]: a lot of mutex implementations require that
+/// a lock is returned to the operating system on the same thread that initially locked the mutex, failing to uphold this
+/// invariant will cause Undefined Behavior
+/// (see last § in [the nomicon](https://doc.rust-lang.org/nomicon/send-and-sync.html)).
+///
+/// It is **always safe** to implement this trait on a type that is `Send`, but no placeholder impl is provided due to limitations in
+/// coherency. Use the [`FullySend`] wrapper in this situation.
+pub unsafe trait MostlySend {}
+
+#[derive(Debug, Clone, Copy, Default, Eq, PartialEq, Ord, PartialOrd, Hash)]
+pub struct FullySend<T>(pub T);
+
+// SAFETY: a type **fully** send is always mostly send as well.
+unsafe impl<T> MostlySend for FullySend<T> where T: Send {}
+
+unsafe impl<T> MostlySend for RefCell<T> where T: MostlySend {}
+
+unsafe impl<T> MostlySend for Option<T> where T: MostlySend {}
+
+impl<T> FullySend<T> {
+    pub fn into(self) -> T {
+        self.0
+    }
+}
+
+impl<T> From<T> for FullySend<T> {
+    fn from(value: T) -> Self {
+        Self(value)
+    }
+}
+
+#[derive(Debug, Clone, Copy, Default, PartialEq, Eq, PartialOrd, Ord, Hash)]
+struct MostlySendWrapper<T>(T);
+
+impl<T: MostlySend> MostlySendWrapper<T> {
+    /// # Safety
+    ///
+    /// - (P1) Users of this type will never access the type concurrently from multiple threads without synchronization
+    unsafe fn new(t: T) -> Self {
+        Self(t)
+    }
+
+    fn as_ref(&self) -> &T {
+        &self.0
+    }
+
+    fn as_mut(&mut self) -> &mut T {
+        &mut self.0
+    }
+
+    fn into_inner(self) -> T {
+        self.0
+    }
+}
+
+/// # Safety
+///
+/// 1. `T` is [`MostlySend`], so by its safety contract it can be accessed by any thread and all of its operations are available
+///   from any thread.
+/// 2. (P1) of `MostlySendWrapper::new` forces the user to never access the value from multiple threads concurrently.
+unsafe impl<T: MostlySend> Send for MostlySendWrapper<T> {}
+
+/// A wrapper around [`thread_local::ThreadLocal`] that accepts [`MostlySend`] `T`s.
+#[derive(Default)]
+pub struct ThreadLocal<T: MostlySend> {
+    inner: thread_local::ThreadLocal<MostlySendWrapper<T>>,
+    // FIXME: this should be necessary
+    //_no_send: PhantomData<*mut ()>,
+}
+
+impl<T: MostlySend> ThreadLocal<T> {
+    pub fn new() -> Self {
+        Self { inner: thread_local::ThreadLocal::new() }
+    }
+
+    pub fn with_capacity(capacity: usize) -> Self {
+        Self { inner: thread_local::ThreadLocal::with_capacity(capacity) }
+    }
+
+    pub fn clear(&mut self) {
+        self.inner.clear()
+    }
+
+    pub fn get(&self) -> Option<&T> {
+        self.inner.get().map(|t| t.as_ref())
+    }
+
+    pub fn get_or<F>(&self, create: F) -> &T
+    where
+        F: FnOnce() -> T,
+    {
+        self.inner.get_or(|| unsafe { MostlySendWrapper::new(create()) }).as_ref()
+    }
+
+    pub fn get_or_try<F, E>(&self, create: F) -> std::result::Result<&T, E>
+    where
+        F: FnOnce() -> std::result::Result<T, E>,
+    {
+        self.inner
+            .get_or_try(|| unsafe { Ok(MostlySendWrapper::new(create()?)) })
+            .map(MostlySendWrapper::as_ref)
+    }
+
+    pub fn get_or_default(&self) -> &T
+    where
+        T: Default,
+    {
+        self.inner.get_or_default().as_ref()
+    }
+
+    pub fn iter_mut(&mut self) -> IterMut<T> {
+        IterMut(self.inner.iter_mut())
+    }
+}
+
+impl<T: MostlySend> IntoIterator for ThreadLocal<T> {
+    type Item = T;
+
+    type IntoIter = IntoIter<T>;
+
+    fn into_iter(self) -> Self::IntoIter {
+        IntoIter(self.inner.into_iter())
+    }
+}
+
+pub struct IterMut<'a, T: MostlySend>(thread_local::IterMut<'a, MostlySendWrapper<T>>);
+
+impl<'a, T: MostlySend> Iterator for IterMut<'a, T> {
+    type Item = &'a mut T;
+
+    fn next(&mut self) -> Option<Self::Item> {
+        self.0.next().map(|t| t.as_mut())
+    }
+}
+
+pub struct IntoIter<T: MostlySend>(thread_local::IntoIter<MostlySendWrapper<T>>);
+
+impl<T: MostlySend> Iterator for IntoIter<T> {
+    type Item = T;
+
+    fn next(&mut self) -> Option<Self::Item> {
+        self.0.next().map(|t| t.into_inner())
+    }
+}

crates/milli/src/vector/error.rs

@@ -60,7 +60,7 @@ pub enum EmbedErrorKind {
     ManualEmbed(String),
     #[error("model not found. Meilisearch will not automatically download models from the Ollama library, please pull the model manually{}", option_info(.0.as_deref(), "server replied with "))]
     OllamaModelNotFoundError(Option<String>),
-    #[error("error deserialization the response body as JSON:\n  - {0}")]
+    #[error("error deserializing the response body as JSON:\n  - {0}")]
     RestResponseDeserialization(std::io::Error),
     #[error("expected a response containing {0} embeddings, got only {1}")]
     RestResponseEmbeddingCount(usize, usize),

crates/milli/src/vector/mod.rs

@@ -1,5 +1,6 @@
 use std::collections::HashMap;
 use std::sync::Arc;
+use std::time::Instant;
 
 use arroy::distances::{BinaryQuantizedCosine, Cosine};
 use arroy::ItemId;
@@ -595,18 +596,26 @@ impl Embedder {
     /// Embed one or multiple texts.
     ///
     /// Each text can be embedded as one or multiple embeddings.
-    pub fn embed(&self, texts: Vec<String>) -> std::result::Result<Vec<Embedding>, EmbedError> {
+    pub fn embed(
+        &self,
+        texts: Vec<String>,
+        deadline: Option<Instant>,
+    ) -> std::result::Result<Vec<Embedding>, EmbedError> {
         match self {
             Embedder::HuggingFace(embedder) => embedder.embed(texts),
-            Embedder::OpenAi(embedder) => embedder.embed(&texts),
-            Embedder::Ollama(embedder) => embedder.embed(&texts),
+            Embedder::OpenAi(embedder) => embedder.embed(&texts, deadline),
+            Embedder::Ollama(embedder) => embedder.embed(&texts, deadline),
             Embedder::UserProvided(embedder) => embedder.embed(&texts),
-            Embedder::Rest(embedder) => embedder.embed(texts),
+            Embedder::Rest(embedder) => embedder.embed(texts, deadline),
         }
     }
 
-    pub fn embed_one(&self, text: String) -> std::result::Result<Embedding, EmbedError> {
-        let mut embedding = self.embed(vec![text])?;
+    pub fn embed_one(
+        &self,
+        text: String,
+        deadline: Option<Instant>,
+    ) -> std::result::Result<Embedding, EmbedError> {
+        let mut embedding = self.embed(vec![text], deadline)?;
         let embedding = embedding.pop().ok_or_else(EmbedError::missing_embedding)?;
         Ok(embedding)
     }

crates/milli/src/vector/ollama.rs

@@ -1,3 +1,5 @@
+use std::time::Instant;
+
 use rayon::iter::{IntoParallelIterator as _, ParallelIterator as _};
 use rayon::slice::ParallelSlice as _;
 
@@ -80,8 +82,9 @@ impl Embedder {
     pub fn embed<S: AsRef<str> + serde::Serialize>(
         &self,
         texts: &[S],
+        deadline: Option<Instant>,
     ) -> Result<Vec<Embedding>, EmbedError> {
-        match self.rest_embedder.embed_ref(texts) {
+        match self.rest_embedder.embed_ref(texts, deadline) {
             Ok(embeddings) => Ok(embeddings),
             Err(EmbedError { kind: EmbedErrorKind::RestOtherStatusCode(404, error), fault: _ }) => {
                 Err(EmbedError::ollama_model_not_found(error))
@@ -97,7 +100,7 @@ impl Embedder {
     ) -> Result<Vec<Vec<Embedding>>, EmbedError> {
         threads
             .install(move || {
-                text_chunks.into_par_iter().map(move |chunk| self.embed(&chunk)).collect()
+                text_chunks.into_par_iter().map(move |chunk| self.embed(&chunk, None)).collect()
             })
             .map_err(|error| EmbedError {
                 kind: EmbedErrorKind::PanicInThreadPool(error),
@@ -114,7 +117,7 @@ impl Embedder {
             .install(move || {
                 let embeddings: Result<Vec<Vec<Embedding>>, _> = texts
                     .par_chunks(self.prompt_count_in_chunk_hint())
-                    .map(move |chunk| self.embed(chunk))
+                    .map(move |chunk| self.embed(chunk, None))
                     .collect();
 
                 let embeddings = embeddings?;

crates/milli/src/vector/openai.rs

@@ -1,3 +1,5 @@
+use std::time::Instant;
+
 use ordered_float::OrderedFloat;
 use rayon::iter::{IntoParallelIterator, ParallelIterator as _};
 use rayon::slice::ParallelSlice as _;
@@ -211,18 +213,23 @@ impl Embedder {
     pub fn embed<S: AsRef<str> + serde::Serialize>(
         &self,
         texts: &[S],
+        deadline: Option<Instant>,
     ) -> Result<Vec<Embedding>, EmbedError> {
-        match self.rest_embedder.embed_ref(texts) {
+        match self.rest_embedder.embed_ref(texts, deadline) {
             Ok(embeddings) => Ok(embeddings),
             Err(EmbedError { kind: EmbedErrorKind::RestBadRequest(error, _), fault: _ }) => {
                 tracing::warn!(error=?error, "OpenAI: received `BAD_REQUEST`. Input was maybe too long, retrying on tokenized version. For best performance, limit the size of your document template.");
-                self.try_embed_tokenized(texts)
+                self.try_embed_tokenized(texts, deadline)
             }
             Err(error) => Err(error),
         }
     }
 
-    fn try_embed_tokenized<S: AsRef<str>>(&self, text: &[S]) -> Result<Vec<Embedding>, EmbedError> {
+    fn try_embed_tokenized<S: AsRef<str>>(
+        &self,
+        text: &[S],
+        deadline: Option<Instant>,
+    ) -> Result<Vec<Embedding>, EmbedError> {
         let mut all_embeddings = Vec::with_capacity(text.len());
         for text in text {
             let text = text.as_ref();
@@ -230,13 +237,13 @@ impl Embedder {
             let encoded = self.tokenizer.encode_ordinary(text);
             let len = encoded.len();
             if len < max_token_count {
-                all_embeddings.append(&mut self.rest_embedder.embed_ref(&[text])?);
+                all_embeddings.append(&mut self.rest_embedder.embed_ref(&[text], deadline)?);
                 continue;
             }
 
             let tokens = &encoded.as_slice()[0..max_token_count];
 
-            let embedding = self.rest_embedder.embed_tokens(tokens)?;
+            let embedding = self.rest_embedder.embed_tokens(tokens, deadline)?;
 
             all_embeddings.push(embedding);
         }
@@ -250,7 +257,7 @@ impl Embedder {
     ) -> Result<Vec<Vec<Embedding>>, EmbedError> {
         threads
             .install(move || {
-                text_chunks.into_par_iter().map(move |chunk| self.embed(&chunk)).collect()
+                text_chunks.into_par_iter().map(move |chunk| self.embed(&chunk, None)).collect()
             })
             .map_err(|error| EmbedError {
                 kind: EmbedErrorKind::PanicInThreadPool(error),
@@ -267,7 +274,7 @@ impl Embedder {
             .install(move || {
                 let embeddings: Result<Vec<Vec<Embedding>>, _> = texts
                     .par_chunks(self.prompt_count_in_chunk_hint())
-                    .map(move |chunk| self.embed(chunk))
+                    .map(move |chunk| self.embed(chunk, None))
                     .collect();
 
                 let embeddings = embeddings?;

crates/milli/src/vector/rest.rs

@@ -1,4 +1,5 @@
 use std::collections::BTreeMap;
+use std::time::Instant;
 
 use deserr::Deserr;
 use rand::Rng;
@@ -153,19 +154,31 @@ impl Embedder {
         Ok(Self { data, dimensions, distribution: options.distribution })
     }
 
-    pub fn embed(&self, texts: Vec<String>) -> Result<Vec<Embedding>, EmbedError> {
-        embed(&self.data, texts.as_slice(), texts.len(), Some(self.dimensions))
+    pub fn embed(
+        &self,
+        texts: Vec<String>,
+        deadline: Option<Instant>,
+    ) -> Result<Vec<Embedding>, EmbedError> {
+        embed(&self.data, texts.as_slice(), texts.len(), Some(self.dimensions), deadline)
     }
 
-    pub fn embed_ref<S>(&self, texts: &[S]) -> Result<Vec<Embedding>, EmbedError>
+    pub fn embed_ref<S>(
+        &self,
+        texts: &[S],
+        deadline: Option<Instant>,
+    ) -> Result<Vec<Embedding>, EmbedError>
     where
         S: AsRef<str> + Serialize,
     {
-        embed(&self.data, texts, texts.len(), Some(self.dimensions))
+        embed(&self.data, texts, texts.len(), Some(self.dimensions), deadline)
     }
 
-    pub fn embed_tokens(&self, tokens: &[usize]) -> Result<Embedding, EmbedError> {
-        let mut embeddings = embed(&self.data, tokens, 1, Some(self.dimensions))?;
+    pub fn embed_tokens(
+        &self,
+        tokens: &[usize],
+        deadline: Option<Instant>,
+    ) -> Result<Embedding, EmbedError> {
+        let mut embeddings = embed(&self.data, tokens, 1, Some(self.dimensions), deadline)?;
         // unwrap: guaranteed that embeddings.len() == 1, otherwise the previous line terminated in error
         Ok(embeddings.pop().unwrap())
     }
@@ -177,7 +190,7 @@ impl Embedder {
     ) -> Result<Vec<Vec<Embedding>>, EmbedError> {
         threads
             .install(move || {
-                text_chunks.into_par_iter().map(move |chunk| self.embed(chunk)).collect()
+                text_chunks.into_par_iter().map(move |chunk| self.embed(chunk, None)).collect()
             })
             .map_err(|error| EmbedError {
                 kind: EmbedErrorKind::PanicInThreadPool(error),
@@ -194,7 +207,7 @@ impl Embedder {
             .install(move || {
                 let embeddings: Result<Vec<Vec<Embedding>>, _> = texts
                     .par_chunks(self.prompt_count_in_chunk_hint())
-                    .map(move |chunk| self.embed_ref(chunk))
+                    .map(move |chunk| self.embed_ref(chunk, None))
                     .collect();
 
                 let embeddings = embeddings?;
@@ -227,7 +240,7 @@ impl Embedder {
 }
 
 fn infer_dimensions(data: &EmbedderData) -> Result<usize, NewEmbedderError> {
-    let v = embed(data, ["test"].as_slice(), 1, None)
+    let v = embed(data, ["test"].as_slice(), 1, None, None)
         .map_err(NewEmbedderError::could_not_determine_dimension)?;
     // unwrap: guaranteed that v.len() == 1, otherwise the previous line terminated in error
     Ok(v.first().unwrap().len())
@@ -238,6 +251,7 @@ fn embed<S>(
     inputs: &[S],
     expected_count: usize,
     expected_dimension: Option<usize>,
+    deadline: Option<Instant>,
 ) -> Result<Vec<Embedding>, EmbedError>
 where
     S: Serialize,
@@ -257,16 +271,27 @@ where
 
     for attempt in 0..10 {
         let response = request.clone().send_json(&body);
-        let result = check_response(response, data.configuration_source);
+        let result = check_response(response, data.configuration_source).and_then(|response| {
+            response_to_embedding(response, data, expected_count, expected_dimension)
+        });
 
         let retry_duration = match result {
-            Ok(response) => {
-                return response_to_embedding(response, data, expected_count, expected_dimension)
-            }
+            Ok(response) => return Ok(response),
             Err(retry) => {
                 tracing::warn!("Failed: {}", retry.error);
+                if let Some(deadline) = deadline {
+                    let now = std::time::Instant::now();
+                    if now > deadline {
+                        tracing::warn!("Could not embed due to deadline");
+                        return Err(retry.into_error());
+                    }
+
+                    let duration_to_deadline = deadline - now;
+                    retry.into_duration(attempt).map(|duration| duration.min(duration_to_deadline))
+                } else {
                     retry.into_duration(attempt)
                 }
+            }
         }?;
 
         let retry_duration = retry_duration.min(std::time::Duration::from_secs(60)); // don't wait more than a minute
@@ -283,6 +308,7 @@ where
     let result = check_response(response, data.configuration_source);
     result.map_err(Retry::into_error).and_then(|response| {
         response_to_embedding(response, data, expected_count, expected_dimension)
+            .map_err(Retry::into_error)
     })
 }
 
@@ -324,20 +350,28 @@ fn response_to_embedding(
     data: &EmbedderData,
     expected_count: usize,
     expected_dimensions: Option<usize>,
-) -> Result<Vec<Embedding>, EmbedError> {
-    let response: serde_json::Value =
-        response.into_json().map_err(EmbedError::rest_response_deserialization)?;
+) -> Result<Vec<Embedding>, Retry> {
+    let response: serde_json::Value = response
+        .into_json()
+        .map_err(EmbedError::rest_response_deserialization)
+        .map_err(Retry::retry_later)?;
 
-    let embeddings = data.response.extract_embeddings(response)?;
+    let embeddings = data.response.extract_embeddings(response).map_err(Retry::give_up)?;
 
     if embeddings.len() != expected_count {
-        return Err(EmbedError::rest_response_embedding_count(expected_count, embeddings.len()));
+        return Err(Retry::give_up(EmbedError::rest_response_embedding_count(
+            expected_count,
+            embeddings.len(),
+        )));
     }
 
     if let Some(dimensions) = expected_dimensions {
         for embedding in &embeddings {
             if embedding.len() != dimensions {
-                return Err(EmbedError::rest_unexpected_dimension(dimensions, embedding.len()));
+                return Err(Retry::give_up(EmbedError::rest_unexpected_dimension(
+                    dimensions,
+                    embedding.len(),
+                )));
             }
         }
     }