Add debugs

5141: Use the right amount of max memory and not impact the settings r=curquiza a=Kerollmops

Fixes #5132. Related to #5125.

Co-authored-by: Kerollmops <clement@meilisearch.com>

Cargo.lock

@@ -492,7 +492,7 @@ checksum = "72b3254f16251a8381aa12e40e3c4d2f0199f8c6508fbecb9d91f575e0fbb8c6"
 [[package]]
 name = "bbqueue"
 version = "0.5.1"
-source = "git+https://github.com/kerollmops/bbqueue#cbb87cc707b5af415ef203bdaf2443e06ba0d6d4"
+source = "git+https://github.com/meilisearch/bbqueue#cbb87cc707b5af415ef203bdaf2443e06ba0d6d4"
 
 [[package]]
 name = "benchmarks"

crates/index-scheduler/src/batch.rs

@@ -29,7 +29,6 @@ use bumpalo::collections::CollectIn;
 use bumpalo::Bump;
 use dump::IndexMetadata;
 use meilisearch_types::batches::BatchId;
-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;
@@ -497,7 +496,6 @@ impl IndexScheduler {
         // 5. We make a batch from the unprioritised tasks. Start by taking the next enqueued task.
         let task_id = if let Some(task_id) = enqueued.min() { task_id } else { return Ok(None) };
         let mut task = self.get_task(rtxn, task_id)?.ok_or(Error::CorruptedTaskQueue)?;
-        current_batch.processing(Some(&mut task));
 
         // If the task is not associated with any index, verify that it is an index swap and
         // create the batch directly. Otherwise, get the index name associated with the task
@@ -507,6 +505,7 @@ impl IndexScheduler {
             index_name
         } else {
             assert!(matches!(&task.kind, KindWithContent::IndexSwap { swaps } if swaps.is_empty()));
+            current_batch.processing(Some(&mut task));
             return Ok(Some((Batch::IndexSwap { task }, current_batch)));
         };
 
@@ -689,7 +688,9 @@ impl IndexScheduler {
                     let index = self.index_mapper.index(&rtxn, name)?;
                     let dst = temp_snapshot_dir.path().join("indexes").join(uuid.to_string());
                     fs::create_dir_all(&dst)?;
-                    index.copy_to_file(dst.join("data.mdb"), CompactionOption::Enabled)?;
+                    index
+                        .copy_to_file(dst.join("data.mdb"), CompactionOption::Enabled)
+                        .map_err(|e| Error::from_milli(e, Some(name.to_string())))?;
                 }
 
                 drop(rtxn);
@@ -791,16 +792,19 @@ impl IndexScheduler {
                             let content_file = self.file_store.get_update(content_file)?;
 
                             let reader = DocumentsBatchReader::from_reader(content_file)
-                                .map_err(milli::Error::from)?;
+                                .map_err(|e| Error::from_milli(e.into(), None))?;
 
                             let (mut cursor, documents_batch_index) =
                                 reader.into_cursor_and_fields_index();
 
-                            while let Some(doc) =
-                                cursor.next_document().map_err(milli::Error::from)?
+                            while let Some(doc) = cursor
+                                .next_document()
+                                .map_err(|e| Error::from_milli(e.into(), None))?
                             {
-                                dump_content_file
-                                    .push_document(&obkv_to_object(doc, &documents_batch_index)?)?;
+                                dump_content_file.push_document(
+                                    &obkv_to_object(doc, &documents_batch_index)
+                                        .map_err(|e| Error::from_milli(e, None))?,
+                                )?;
                             }
                             dump_content_file.flush()?;
                         }
@@ -814,27 +818,41 @@ impl IndexScheduler {
                     let metadata = IndexMetadata {
                         uid: uid.to_owned(),
                         primary_key: index.primary_key(&rtxn)?.map(String::from),
-                        created_at: index.created_at(&rtxn)?,
-                        updated_at: index.updated_at(&rtxn)?,
+                        created_at: index
+                            .created_at(&rtxn)
+                            .map_err(|e| Error::from_milli(e, Some(uid.to_string())))?,
+                        updated_at: index
+                            .updated_at(&rtxn)
+                            .map_err(|e| Error::from_milli(e, Some(uid.to_string())))?,
                     };
                     let mut index_dumper = dump.create_index(uid, &metadata)?;
 
                     let fields_ids_map = index.fields_ids_map(&rtxn)?;
                     let all_fields: Vec<_> = fields_ids_map.iter().map(|(id, _)| id).collect();
-                    let embedding_configs = index.embedding_configs(&rtxn)?;
+                    let embedding_configs = index
+                        .embedding_configs(&rtxn)
+                        .map_err(|e| Error::from_milli(e, Some(uid.to_string())))?;
 
+                    let documents = index
+                        .all_documents(&rtxn)
+                        .map_err(|e| Error::from_milli(e, Some(uid.to_string())))?;
                     // 3.1. Dump the documents
-                    for ret in index.all_documents(&rtxn)? {
+                    for ret in documents {
                         if self.must_stop_processing.get() {
                             return Err(Error::AbortedTask);
                         }
 
-                        let (id, doc) = ret?;
+                        let (id, doc) =
+                            ret.map_err(|e| Error::from_milli(e, Some(uid.to_string())))?;
 
-                        let mut document = milli::obkv_to_json(&all_fields, &fields_ids_map, doc)?;
+                        let mut document =
+                            milli::obkv_to_json(&all_fields, &fields_ids_map, doc)
+                                .map_err(|e| Error::from_milli(e, Some(uid.to_string())))?;
 
                         'inject_vectors: {
-                            let embeddings = index.embeddings(&rtxn, id)?;
+                            let embeddings = index
+                                .embeddings(&rtxn, id)
+                                .map_err(|e| Error::from_milli(e, Some(uid.to_string())))?;
 
                             if embeddings.is_empty() {
                                 break 'inject_vectors;
@@ -845,7 +863,7 @@ impl IndexScheduler {
                                 .or_insert(serde_json::Value::Object(Default::default()));
 
                             let serde_json::Value::Object(vectors) = vectors else {
-                                return Err(milli::Error::UserError(
+                                let user_err = milli::Error::UserError(
                                     milli::UserError::InvalidVectorsMapType {
                                         document_id: {
                                             if let Ok(Some(Ok(index))) = index
@@ -859,8 +877,9 @@ impl IndexScheduler {
                                         },
                                         value: vectors.clone(),
                                     },
-                                )
-                                .into());
+                                );
+
+                                return Err(Error::from_milli(user_err, Some(uid.to_string())));
                             };
 
                             for (embedder_name, embeddings) in embeddings {
@@ -890,7 +909,8 @@ impl IndexScheduler {
                         index,
                         &rtxn,
                         meilisearch_types::settings::SecretPolicy::RevealSecrets,
-                    )?;
+                    )
+                    .map_err(|e| Error::from_milli(e, Some(uid.to_string())))?;
                     index_dumper.settings(&settings)?;
                     Ok(())
                 })?;
@@ -946,7 +966,8 @@ impl IndexScheduler {
                 // the entire batch.
                 let res = || -> Result<()> {
                     let index_rtxn = index.read_txn()?;
-                    let stats = crate::index_mapper::IndexStats::new(&index, &index_rtxn)?;
+                    let stats = crate::index_mapper::IndexStats::new(&index, &index_rtxn)
+                        .map_err(|e| Error::from_milli(e, Some(index_uid.to_string())))?;
                     let mut wtxn = self.env.write_txn()?;
                     self.index_mapper.store_stats_of(&mut wtxn, &index_uid, &stats)?;
                     wtxn.commit()?;
@@ -988,10 +1009,12 @@ impl IndexScheduler {
                     );
                     builder.set_primary_key(primary_key);
                     let must_stop_processing = self.must_stop_processing.clone();
-                    builder.execute(
-                        |indexing_step| tracing::debug!(update = ?indexing_step),
-                        || must_stop_processing.get(),
-                    )?;
+                    builder
+                        .execute(
+                            |indexing_step| tracing::debug!(update = ?indexing_step),
+                            || must_stop_processing.get(),
+                        )
+                        .map_err(|e| Error::from_milli(e, Some(index_uid.to_string())))?;
                     index_wtxn.commit()?;
                 }
 
@@ -1008,7 +1031,8 @@ impl IndexScheduler {
                 let res = || -> Result<()> {
                     let mut wtxn = self.env.write_txn()?;
                     let index_rtxn = index.read_txn()?;
-                    let stats = crate::index_mapper::IndexStats::new(&index, &index_rtxn)?;
+                    let stats = crate::index_mapper::IndexStats::new(&index, &index_rtxn)
+                        .map_err(|e| Error::from_milli(e, Some(index_uid.clone())))?;
                     self.index_mapper.store_stats_of(&mut wtxn, &index_uid, &stats)?;
                     wtxn.commit()?;
                     Ok(())
@@ -1031,7 +1055,9 @@ impl IndexScheduler {
                 let number_of_documents = || -> Result<u64> {
                     let index = self.index_mapper.index(&wtxn, &index_uid)?;
                     let index_rtxn = index.read_txn()?;
-                    Ok(index.number_of_documents(&index_rtxn)?)
+                    index
+                        .number_of_documents(&index_rtxn)
+                        .map_err(|e| Error::from_milli(e, Some(index_uid.to_string())))
                 }()
                 .unwrap_or_default();
 
@@ -1188,8 +1214,10 @@ impl IndexScheduler {
         };
 
         match operation {
-            IndexOperation::DocumentClear { mut tasks, .. } => {
-                let count = milli::update::ClearDocuments::new(index_wtxn, index).execute()?;
+            IndexOperation::DocumentClear { index_uid, mut tasks } => {
+                let count = milli::update::ClearDocuments::new(index_wtxn, index)
+                    .execute()
+                    .map_err(|e| Error::from_milli(e, Some(index_uid)))?;
 
                 let mut first_clear_found = false;
                 for task in &mut tasks {
@@ -1209,7 +1237,7 @@ impl IndexScheduler {
                 Ok(tasks)
             }
             IndexOperation::DocumentOperation {
-                index_uid: _,
+                index_uid,
                 primary_key,
                 method,
                 operations,
@@ -1235,13 +1263,17 @@ impl IndexScheduler {
 
                 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)?;
+                let embedders = index
+                    .embedding_configs(index_wtxn)
+                    .map_err(|e| Error::from_milli(e, Some(index_uid.clone())))?;
+                let embedders = self.embedders(index_uid.clone(), embedders)?;
                 for operation in operations {
                     match operation {
                         DocumentOperation::Add(_content_uuid) => {
                             let mmap = content_files_iter.next().unwrap();
-                            indexer.add_documents(mmap)?;
+                            indexer
+                                .add_documents(mmap)
+                                .map_err(|e| Error::from_milli(e, Some(index_uid.clone())))?;
                         }
                         DocumentOperation::Delete(document_ids) => {
                             let document_ids: bumpalo::collections::vec::Vec<_> = document_ids
@@ -1266,15 +1298,17 @@ impl IndexScheduler {
                     }
                 };
 
-                let (document_changes, operation_stats, primary_key) = indexer.into_changes(
-                    &indexer_alloc,
-                    index,
-                    &rtxn,
-                    primary_key.as_deref(),
-                    &mut new_fields_ids_map,
-                    &|| must_stop_processing.get(),
-                    &send_progress,
-                )?;
+                let (document_changes, operation_stats, primary_key) = indexer
+                    .into_changes(
+                        &indexer_alloc,
+                        index,
+                        &rtxn,
+                        primary_key.as_deref(),
+                        &mut new_fields_ids_map,
+                        &|| must_stop_processing.get(),
+                        &send_progress,
+                    )
+                    .map_err(|e| Error::from_milli(e, Some(index_uid.clone())))?;
 
                 let mut addition = 0;
                 for (stats, task) in operation_stats.into_iter().zip(&mut tasks) {
@@ -1321,14 +1355,15 @@ impl IndexScheduler {
                         embedders,
                         &|| must_stop_processing.get(),
                         &send_progress,
-                    )?;
+                    )
+                    .map_err(|e| Error::from_milli(e, Some(index_uid.clone())))?;
 
                     tracing::info!(indexing_result = ?addition, processed_in = ?started_processing_at.elapsed(), "document indexing done");
                 }
 
                 Ok(tasks)
             }
-            IndexOperation::DocumentEdition { mut task, .. } => {
+            IndexOperation::DocumentEdition { index_uid, mut task } => {
                 let (filter, code) = if let KindWithContent::DocumentEdition {
                     filter_expr,
                     context: _,
@@ -1342,16 +1377,11 @@ impl IndexScheduler {
                 };
 
                 let candidates = match filter.as_ref().map(Filter::from_json) {
-                    Some(Ok(Some(filter))) => {
-                        filter.evaluate(index_wtxn, index).map_err(|err| match err {
-                            milli::Error::UserError(milli::UserError::InvalidFilter(_)) => {
-                                Error::from(err).with_custom_error_code(Code::InvalidDocumentFilter)
-                            }
-                            e => e.into(),
-                        })?
-                    }
+                    Some(Ok(Some(filter))) => filter
+                        .evaluate(index_wtxn, index)
+                        .map_err(|err| Error::from_milli(err, Some(index_uid.clone())))?,
                     None | Some(Ok(None)) => index.documents_ids(index_wtxn)?,
-                    Some(Err(e)) => return Err(e.into()),
+                    Some(Err(e)) => return Err(Error::from_milli(e, Some(index_uid.clone()))),
                 };
 
                 let (original_filter, context, function) = if let Some(Details::DocumentEdition {
@@ -1386,8 +1416,9 @@ impl IndexScheduler {
                 // candidates not empty => index not empty => a primary key is set
                 let primary_key = index.primary_key(&rtxn)?.unwrap();
 
-                let primary_key = PrimaryKey::new_or_insert(primary_key, &mut new_fields_ids_map)
-                    .map_err(milli::Error::from)?;
+                let primary_key =
+                    PrimaryKey::new_or_insert(primary_key, &mut new_fields_ids_map)
+                        .map_err(|err| Error::from_milli(err.into(), Some(index_uid.clone())))?;
 
                 let result_count = Ok((candidates.len(), candidates.len())) as Result<_>;
 
@@ -1406,11 +1437,17 @@ impl IndexScheduler {
                     };
 
                     let indexer = UpdateByFunction::new(candidates, context.clone(), code.clone());
-                    let document_changes =
-                        pool.install(|| indexer.into_changes(&primary_key)).unwrap()?;
-
-                    let embedders = index.embedding_configs(index_wtxn)?;
-                    let embedders = self.embedders(embedders)?;
+                    let document_changes = pool
+                        .install(|| {
+                            indexer
+                                .into_changes(&primary_key)
+                                .map_err(|err| Error::from_milli(err, Some(index_uid.clone())))
+                        })
+                        .unwrap()?;
+                    let embedders = index
+                        .embedding_configs(index_wtxn)
+                        .map_err(|err| Error::from_milli(err, Some(index_uid.clone())))?;
+                    let embedders = self.embedders(index_uid.clone(), embedders)?;
 
                     indexer::index(
                         index_wtxn,
@@ -1424,7 +1461,8 @@ impl IndexScheduler {
                         embedders,
                         &|| must_stop_processing.get(),
                         &send_progress,
-                    )?;
+                    )
+                    .map_err(|err| Error::from_milli(err, Some(index_uid.clone())))?;
 
                     // tracing::info!(indexing_result = ?addition, processed_in = ?started_processing_at.elapsed(), "document indexing done");
                 }
@@ -1455,7 +1493,7 @@ impl IndexScheduler {
 
                 Ok(vec![task])
             }
-            IndexOperation::DocumentDeletion { mut tasks, index_uid: _ } => {
+            IndexOperation::DocumentDeletion { mut tasks, index_uid } => {
                 let mut to_delete = RoaringBitmap::new();
                 let external_documents_ids = index.external_documents_ids();
 
@@ -1476,35 +1514,23 @@ impl IndexScheduler {
                                 deleted_documents: Some(will_be_removed),
                             });
                         }
-                        KindWithContent::DocumentDeletionByFilter { index_uid: _, filter_expr } => {
+                        KindWithContent::DocumentDeletionByFilter { index_uid, filter_expr } => {
                             let before = to_delete.len();
                             let filter = match Filter::from_json(filter_expr) {
                                 Ok(filter) => filter,
                                 Err(err) => {
                                     // theorically, this should be catched by deserr before reaching the index-scheduler and cannot happens
                                     task.status = Status::Failed;
-                                    task.error = match err {
-                                        milli::Error::UserError(
-                                            milli::UserError::InvalidFilterExpression { .. },
-                                        ) => Some(
-                                            Error::from(err)
-                                                .with_custom_error_code(Code::InvalidDocumentFilter)
-                                                .into(),
-                                        ),
-                                        e => Some(e.into()),
-                                    };
+                                    task.error = Some(
+                                        Error::from_milli(err, Some(index_uid.clone())).into(),
+                                    );
                                     None
                                 }
                             };
                             if let Some(filter) = filter {
-                                let candidates =
-                                    filter.evaluate(index_wtxn, index).map_err(|err| match err {
-                                        milli::Error::UserError(
-                                            milli::UserError::InvalidFilter(_),
-                                        ) => Error::from(err)
-                                            .with_custom_error_code(Code::InvalidDocumentFilter),
-                                        e => e.into(),
-                                    });
+                                let candidates = filter
+                                    .evaluate(index_wtxn, index)
+                                    .map_err(|err| Error::from_milli(err, Some(index_uid.clone())));
                                 match candidates {
                                     Ok(candidates) => to_delete |= candidates,
                                     Err(err) => {
@@ -1540,8 +1566,9 @@ impl IndexScheduler {
                 // to_delete not empty => index not empty => primary key set
                 let primary_key = index.primary_key(&rtxn)?.unwrap();
 
-                let primary_key = PrimaryKey::new_or_insert(primary_key, &mut new_fields_ids_map)
-                    .map_err(milli::Error::from)?;
+                let primary_key =
+                    PrimaryKey::new_or_insert(primary_key, &mut new_fields_ids_map)
+                        .map_err(|err| Error::from_milli(err.into(), Some(index_uid.clone())))?;
 
                 if !tasks.iter().all(|res| res.error.is_some()) {
                     let local_pool;
@@ -1560,8 +1587,10 @@ impl IndexScheduler {
                     let mut indexer = indexer::DocumentDeletion::new();
                     indexer.delete_documents_by_docids(to_delete);
                     let document_changes = indexer.into_changes(&indexer_alloc, primary_key);
-                    let embedders = index.embedding_configs(index_wtxn)?;
-                    let embedders = self.embedders(embedders)?;
+                    let embedders = index
+                        .embedding_configs(index_wtxn)
+                        .map_err(|err| Error::from_milli(err, Some(index_uid.clone())))?;
+                    let embedders = self.embedders(index_uid.clone(), embedders)?;
 
                     indexer::index(
                         index_wtxn,
@@ -1575,14 +1604,15 @@ impl IndexScheduler {
                         embedders,
                         &|| must_stop_processing.get(),
                         &send_progress,
-                    )?;
+                    )
+                    .map_err(|err| Error::from_milli(err, Some(index_uid.clone())))?;
 
                     // tracing::info!(indexing_result = ?addition, processed_in = ?started_processing_at.elapsed(), "document indexing done");
                 }
 
                 Ok(tasks)
             }
-            IndexOperation::Settings { index_uid: _, settings, mut tasks } => {
+            IndexOperation::Settings { index_uid, settings, mut tasks } => {
                 let indexer_config = self.index_mapper.indexer_config();
                 let mut builder = milli::update::Settings::new(index_wtxn, index, indexer_config);
 
@@ -1596,10 +1626,12 @@ impl IndexScheduler {
                     task.status = Status::Succeeded;
                 }
 
-                builder.execute(
-                    |indexing_step| tracing::debug!(update = ?indexing_step),
-                    || must_stop_processing.get(),
-                )?;
+                builder
+                    .execute(
+                        |indexing_step| tracing::debug!(update = ?indexing_step),
+                        || must_stop_processing.get(),
+                    )
+                    .map_err(|err| Error::from_milli(err, Some(index_uid.clone())))?;
 
                 Ok(tasks)
             }

crates/index-scheduler/src/error.rs

@@ -1,13 +1,12 @@
 use std::fmt::Display;
 
+use crate::TaskId;
 use meilisearch_types::batches::BatchId;
 use meilisearch_types::error::{Code, ErrorCode};
 use meilisearch_types::tasks::{Kind, Status};
 use meilisearch_types::{heed, milli};
 use thiserror::Error;
 
-use crate::TaskId;
-
 #[derive(Copy, Clone, Debug, PartialEq, Eq)]
 pub enum DateField {
     BeforeEnqueuedAt,
@@ -122,8 +121,11 @@ pub enum Error {
     Dump(#[from] dump::Error),
     #[error(transparent)]
     Heed(#[from] heed::Error),
-    #[error(transparent)]
-    Milli(#[from] milli::Error),
+    #[error("{}", match .index_uid {
+        Some(uid) if !uid.is_empty() => format!("Index `{}`: {error}", uid),
+        _ => format!("{error}")
+    })]
+    Milli { error: milli::Error, index_uid: Option<String> },
     #[error("An unexpected crash occurred when processing the task.")]
     ProcessBatchPanicked,
     #[error(transparent)]
@@ -190,7 +192,7 @@ impl Error {
             | Error::AbortedTask
             | Error::Dump(_)
             | Error::Heed(_)
-            | Error::Milli(_)
+            | Error::Milli { .. }
             | Error::ProcessBatchPanicked
             | Error::FileStore(_)
             | Error::IoError(_)
@@ -209,6 +211,20 @@ impl Error {
     pub fn with_custom_error_code(self, code: Code) -> Self {
         Self::WithCustomErrorCode(code, Box::new(self))
     }
+
+    pub fn from_milli(err: milli::Error, index_uid: Option<String>) -> Self {
+        match err {
+            milli::Error::UserError(milli::UserError::InvalidFilter(_)) => {
+                Self::Milli { error: err, index_uid }
+                    .with_custom_error_code(Code::InvalidDocumentFilter)
+            }
+            milli::Error::UserError(milli::UserError::InvalidFilterExpression { .. }) => {
+                Self::Milli { error: err, index_uid }
+                    .with_custom_error_code(Code::InvalidDocumentFilter)
+            }
+            _ => Self::Milli { error: err, index_uid },
+        }
+    }
 }
 
 impl ErrorCode for Error {
@@ -236,7 +252,7 @@ impl ErrorCode for Error {
             // TODO: not sure of the Code to use
             Error::NoSpaceLeftInTaskQueue => Code::NoSpaceLeftOnDevice,
             Error::Dump(e) => e.error_code(),
-            Error::Milli(e) => e.error_code(),
+            Error::Milli { error, .. } => error.error_code(),
             Error::ProcessBatchPanicked => Code::Internal,
             Error::Heed(e) => e.error_code(),
             Error::HeedTransaction(e) => e.error_code(),

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

@@ -3,14 +3,13 @@ use std::path::Path;
 use std::time::Duration;
 
 use meilisearch_types::heed::{EnvClosingEvent, EnvFlags, EnvOpenOptions};
-use meilisearch_types::milli::Index;
+use meilisearch_types::milli::{Index, Result};
 use time::OffsetDateTime;
 use uuid::Uuid;
 
 use super::IndexStatus::{self, Available, BeingDeleted, Closing, Missing};
+use crate::clamp_to_page_size;
 use crate::lru::{InsertionOutcome, LruMap};
-use crate::{clamp_to_page_size, Result};
-
 /// Keep an internally consistent view of the open indexes in memory.
 ///
 /// This view is made of an LRU cache that will evict the least frequently used indexes when new indexes are opened.

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

@@ -3,8 +3,13 @@ use std::sync::{Arc, RwLock};
 use std::time::Duration;
 use std::{fs, thread};
 
+use self::index_map::IndexMap;
+use self::IndexStatus::{Available, BeingDeleted, Closing, Missing};
+use crate::uuid_codec::UuidCodec;
+use crate::{Error, Result};
 use meilisearch_types::heed::types::{SerdeJson, Str};
 use meilisearch_types::heed::{Database, Env, RoTxn, RwTxn};
+use meilisearch_types::milli;
 use meilisearch_types::milli::update::IndexerConfig;
 use meilisearch_types::milli::{FieldDistribution, Index};
 use serde::{Deserialize, Serialize};
@@ -12,11 +17,6 @@ use time::OffsetDateTime;
 use tracing::error;
 use uuid::Uuid;
 
-use self::index_map::IndexMap;
-use self::IndexStatus::{Available, BeingDeleted, Closing, Missing};
-use crate::uuid_codec::UuidCodec;
-use crate::{Error, Result};
-
 mod index_map;
 
 const INDEX_MAPPING: &str = "index-mapping";
@@ -121,7 +121,7 @@ impl IndexStats {
     /// # Parameters
     ///
     /// - rtxn: a RO transaction for the index, obtained from `Index::read_txn()`.
-    pub fn new(index: &Index, rtxn: &RoTxn) -> Result<Self> {
+    pub fn new(index: &Index, rtxn: &RoTxn) -> milli::Result<Self> {
         Ok(IndexStats {
             number_of_documents: index.number_of_documents(rtxn)?,
             database_size: index.on_disk_size()?,
@@ -183,13 +183,18 @@ impl IndexMapper {
                 // Error if the UUIDv4 somehow already exists in the map, since it should be fresh.
                 // This is very unlikely to happen in practice.
                 // TODO: it would be better to lazily create the index. But we need an Index::open function for milli.
-                let index = self.index_map.write().unwrap().create(
-                    &uuid,
-                    &index_path,
-                    date,
-                    self.enable_mdb_writemap,
-                    self.index_base_map_size,
-                )?;
+                let index = self
+                    .index_map
+                    .write()
+                    .unwrap()
+                    .create(
+                        &uuid,
+                        &index_path,
+                        date,
+                        self.enable_mdb_writemap,
+                        self.index_base_map_size,
+                    )
+                    .map_err(|e| Error::from_milli(e, Some(uuid.to_string())))?;
 
                 wtxn.commit()?;
 
@@ -357,7 +362,9 @@ impl IndexMapper {
                     };
                     let index_path = self.base_path.join(uuid.to_string());
                     // take the lock to reopen the environment.
-                    reopen.reopen(&mut self.index_map.write().unwrap(), &index_path)?;
+                    reopen
+                        .reopen(&mut self.index_map.write().unwrap(), &index_path)
+                        .map_err(|e| Error::from_milli(e, Some(uuid.to_string())))?;
                     continue;
                 }
                 BeingDeleted => return Err(Error::IndexNotFound(name.to_string())),
@@ -372,13 +379,15 @@ impl IndexMapper {
                         Missing => {
                             let index_path = self.base_path.join(uuid.to_string());
 
-                            break index_map.create(
-                                &uuid,
-                                &index_path,
-                                None,
-                                self.enable_mdb_writemap,
-                                self.index_base_map_size,
-                            )?;
+                            break index_map
+                                .create(
+                                    &uuid,
+                                    &index_path,
+                                    None,
+                                    self.enable_mdb_writemap,
+                                    self.index_base_map_size,
+                                )
+                                .map_err(|e| Error::from_milli(e, Some(uuid.to_string())))?;
                         }
                         Available(index) => break index,
                         Closing(_) => {
@@ -460,6 +469,7 @@ impl IndexMapper {
                 let index = self.index(rtxn, index_uid)?;
                 let index_rtxn = index.read_txn()?;
                 IndexStats::new(&index, &index_rtxn)
+                    .map_err(|e| Error::from_milli(e, Some(uuid.to_string())))
             }
         }
     }

crates/index-scheduler/src/lib.rs

@@ -1440,7 +1440,7 @@ impl IndexScheduler {
 
         // if the task doesn't delete anything and 50% of the task queue is full, we must refuse to enqueue the incomming task
         if !matches!(&kind, KindWithContent::TaskDeletion { tasks, .. } if !tasks.is_empty())
-            && (self.env.non_free_pages_size()? * 100) / self.env.info().map_size as u64 > 50
+            && (self.env.non_free_pages_size()? * 100) / self.env.info().map_size as u64 > 40
         {
             return Err(Error::NoSpaceLeftInTaskQueue);
         }
@@ -1678,9 +1678,10 @@ impl IndexScheduler {
                 tracing::info!("A batch of tasks was successfully completed with {success} successful tasks and {failure} failed tasks.");
             }
             // If we have an abortion error we must stop the tick here and re-schedule tasks.
-            Err(Error::Milli(milli::Error::InternalError(
-                milli::InternalError::AbortedIndexation,
-            )))
+            Err(Error::Milli {
+                error: milli::Error::InternalError(milli::InternalError::AbortedIndexation),
+                ..
+            })
             | Err(Error::AbortedTask) => {
                 #[cfg(test)]
                 self.breakpoint(Breakpoint::AbortedIndexation);
@@ -1699,9 +1700,10 @@ impl IndexScheduler {
             // 2. close the associated environment
             // 3. resize it
             // 4. re-schedule tasks
-            Err(Error::Milli(milli::Error::UserError(
-                milli::UserError::MaxDatabaseSizeReached,
-            ))) if index_uid.is_some() => {
+            Err(Error::Milli {
+                error: milli::Error::UserError(milli::UserError::MaxDatabaseSizeReached),
+                ..
+            }) if index_uid.is_some() => {
                 // fixme: add index_uid to match to avoid the unwrap
                 let index_uid = index_uid.unwrap();
                 // fixme: handle error more gracefully? not sure when this could happen
@@ -1943,6 +1945,7 @@ impl IndexScheduler {
     // TODO: consider using a type alias or a struct embedder/template
     pub fn embedders(
         &self,
+        index_uid: String,
         embedding_configs: Vec<IndexEmbeddingConfig>,
     ) -> Result<EmbeddingConfigs> {
         let res: Result<_> = embedding_configs
@@ -1953,8 +1956,12 @@ impl IndexScheduler {
                      config: milli::vector::EmbeddingConfig { embedder_options, prompt, quantized },
                      ..
                  }| {
-                    let prompt =
-                        Arc::new(prompt.try_into().map_err(meilisearch_types::milli::Error::from)?);
+                    let prompt = Arc::new(
+                        prompt
+                            .try_into()
+                            .map_err(meilisearch_types::milli::Error::from)
+                            .map_err(|err| Error::from_milli(err, Some(index_uid.clone())))?,
+                    );
                     // optimistically return existing embedder
                     {
                         let embedders = self.embedders.read().unwrap();
@@ -1970,7 +1977,9 @@ impl IndexScheduler {
                     let embedder = Arc::new(
                         Embedder::new(embedder_options.clone())
                             .map_err(meilisearch_types::milli::vector::Error::from)
-                            .map_err(meilisearch_types::milli::Error::from)?,
+                            .map_err(|err| {
+                                Error::from_milli(err.into(), Some(index_uid.clone()))
+                            })?,
                     );
                     {
                         let mut embedders = self.embedders.write().unwrap();
@@ -4319,10 +4328,35 @@ mod tests {
         let proc = index_scheduler.processing_tasks.read().unwrap().clone();
 
         let query = Query { statuses: Some(vec![Status::Processing]), ..Default::default() };
-        let (batches, _) = index_scheduler
-            .get_batch_ids_from_authorized_indexes(&rtxn, &proc, &query, &AuthFilter::default())
+        let (mut batches, _) = index_scheduler
+            .get_batches_from_authorized_indexes(query.clone(), &AuthFilter::default())
             .unwrap();
-        snapshot!(snapshot_bitmap(&batches), @"[0,]"); // only the processing batch in the first tick
+        assert_eq!(batches.len(), 1);
+        batches[0].started_at = OffsetDateTime::UNIX_EPOCH;
+        // Insta cannot snapshot our batches because the batch stats contains an enum as key: https://github.com/mitsuhiko/insta/issues/689
+        let batch = serde_json::to_string_pretty(&batches[0]).unwrap();
+        snapshot!(batch, @r#"
+        {
+          "uid": 0,
+          "details": {
+            "primaryKey": "mouse"
+          },
+          "stats": {
+            "totalNbTasks": 1,
+            "status": {
+              "processing": 1
+            },
+            "types": {
+              "indexCreation": 1
+            },
+            "indexUids": {
+              "catto": 1
+            }
+          },
+          "startedAt": "1970-01-01T00:00:00Z",
+          "finishedAt": null
+        }
+        "#);
 
         let query = Query { statuses: Some(vec![Status::Enqueued]), ..Default::default() };
         let (batches, _) = index_scheduler
@@ -6146,7 +6180,7 @@ mod tests {
             insta::assert_json_snapshot!(simple_hf_config.embedder_options);
             let simple_hf_name = name.clone();
 
-            let configs = index_scheduler.embedders(configs).unwrap();
+            let configs = index_scheduler.embedders("doggos".to_string(), configs).unwrap();
             let (hf_embedder, _, _) = configs.get(&simple_hf_name).unwrap();
             let beagle_embed =
                 hf_embedder.embed_one(S("Intel the beagle best doggo"), None).unwrap();

crates/index-scheduler/src/snapshots/lib.rs/cancel_mix_of_tasks/aborted_indexation.snap

@@ -5,7 +5,7 @@ snapshot_kind: text
 ### Autobatching Enabled = true
 ### Processing batch Some(1):
 [1,]
-{uid: 1, details: {"receivedDocuments":2,"indexedDocuments":null}, stats: {"totalNbTasks":2,"status":{"enqueued":2},"types":{"documentAdditionOrUpdate":2},"indexUids":{"beavero":2}}, }
+{uid: 1, details: {"receivedDocuments":1,"indexedDocuments":null}, stats: {"totalNbTasks":1,"status":{"processing":1},"types":{"documentAdditionOrUpdate":1},"indexUids":{"beavero":1}}, }
 ----------------------------------------------------------------------
 ### All Tasks:
 0 {uid: 0, batch_uid: 0, status: succeeded, details: { received_documents: 1, indexed_documents: Some(1) }, kind: DocumentAdditionOrUpdate { index_uid: "catto", primary_key: None, method: ReplaceDocuments, content_file: 00000000-0000-0000-0000-000000000000, documents_count: 1, allow_index_creation: true }}

crates/index-scheduler/src/snapshots/lib.rs/cancel_mix_of_tasks/processing_second_task_cancel_enqueued.snap

@@ -5,7 +5,7 @@ snapshot_kind: text
 ### Autobatching Enabled = true
 ### Processing batch Some(1):
 [1,]
-{uid: 1, details: {"receivedDocuments":2,"indexedDocuments":null}, stats: {"totalNbTasks":2,"status":{"enqueued":2},"types":{"documentAdditionOrUpdate":2},"indexUids":{"beavero":2}}, }
+{uid: 1, details: {"receivedDocuments":1,"indexedDocuments":null}, stats: {"totalNbTasks":1,"status":{"processing":1},"types":{"documentAdditionOrUpdate":1},"indexUids":{"beavero":1}}, }
 ----------------------------------------------------------------------
 ### All Tasks:
 0 {uid: 0, batch_uid: 0, status: succeeded, details: { received_documents: 1, indexed_documents: Some(1) }, kind: DocumentAdditionOrUpdate { index_uid: "catto", primary_key: None, method: ReplaceDocuments, content_file: 00000000-0000-0000-0000-000000000000, documents_count: 1, allow_index_creation: true }}

crates/index-scheduler/src/snapshots/lib.rs/cancel_processing_dump/cancel_registered.snap

@@ -5,7 +5,7 @@ snapshot_kind: text
 ### Autobatching Enabled = true
 ### Processing batch Some(0):
 [0,]
-{uid: 0, details: {"dumpUid":null}, stats: {"totalNbTasks":1,"status":{"enqueued":1},"types":{"dumpCreation":1},"indexUids":{}}, }
+{uid: 0, details: {"dumpUid":null}, stats: {"totalNbTasks":1,"status":{"processing":1},"types":{"dumpCreation":1},"indexUids":{}}, }
 ----------------------------------------------------------------------
 ### All Tasks:
 0 {uid: 0, status: enqueued, details: { dump_uid: None }, kind: DumpCreation { keys: [], instance_uid: None }}

crates/index-scheduler/src/snapshots/lib.rs/cancel_processing_task/aborted_indexation.snap

@@ -5,7 +5,7 @@ snapshot_kind: text
 ### Autobatching Enabled = true
 ### Processing batch Some(0):
 [0,]
-{uid: 0, details: {"receivedDocuments":2,"indexedDocuments":null}, stats: {"totalNbTasks":2,"status":{"enqueued":2},"types":{"documentAdditionOrUpdate":2},"indexUids":{"catto":2}}, }
+{uid: 0, details: {"receivedDocuments":1,"indexedDocuments":null}, stats: {"totalNbTasks":1,"status":{"processing":1},"types":{"documentAdditionOrUpdate":1},"indexUids":{"catto":1}}, }
 ----------------------------------------------------------------------
 ### All Tasks:
 0 {uid: 0, status: enqueued, details: { received_documents: 1, indexed_documents: None }, kind: DocumentAdditionOrUpdate { index_uid: "catto", primary_key: None, method: ReplaceDocuments, content_file: 00000000-0000-0000-0000-000000000000, documents_count: 1, allow_index_creation: true }}

crates/index-scheduler/src/snapshots/lib.rs/cancel_processing_task/cancel_task_registered.snap

@@ -5,7 +5,7 @@ snapshot_kind: text
 ### Autobatching Enabled = true
 ### Processing batch Some(0):
 [0,]
-{uid: 0, details: {"receivedDocuments":2,"indexedDocuments":null}, stats: {"totalNbTasks":2,"status":{"enqueued":2},"types":{"documentAdditionOrUpdate":2},"indexUids":{"catto":2}}, }
+{uid: 0, details: {"receivedDocuments":1,"indexedDocuments":null}, stats: {"totalNbTasks":1,"status":{"processing":1},"types":{"documentAdditionOrUpdate":1},"indexUids":{"catto":1}}, }
 ----------------------------------------------------------------------
 ### All Tasks:
 0 {uid: 0, status: enqueued, details: { received_documents: 1, indexed_documents: None }, kind: DocumentAdditionOrUpdate { index_uid: "catto", primary_key: None, method: ReplaceDocuments, content_file: 00000000-0000-0000-0000-000000000000, documents_count: 1, allow_index_creation: true }}

crates/index-scheduler/src/snapshots/lib.rs/cancel_processing_task/initial_task_processing.snap

@@ -5,7 +5,7 @@ snapshot_kind: text
 ### Autobatching Enabled = true
 ### Processing batch Some(0):
 [0,]
-{uid: 0, details: {"receivedDocuments":2,"indexedDocuments":null}, stats: {"totalNbTasks":2,"status":{"enqueued":2},"types":{"documentAdditionOrUpdate":2},"indexUids":{"catto":2}}, }
+{uid: 0, details: {"receivedDocuments":1,"indexedDocuments":null}, stats: {"totalNbTasks":1,"status":{"processing":1},"types":{"documentAdditionOrUpdate":1},"indexUids":{"catto":1}}, }
 ----------------------------------------------------------------------
 ### All Tasks:
 0 {uid: 0, status: enqueued, details: { received_documents: 1, indexed_documents: None }, kind: DocumentAdditionOrUpdate { index_uid: "catto", primary_key: None, method: ReplaceDocuments, content_file: 00000000-0000-0000-0000-000000000000, documents_count: 1, allow_index_creation: true }}

crates/index-scheduler/src/snapshots/lib.rs/document_addition/after_the_batch_creation.snap

@@ -5,7 +5,7 @@ snapshot_kind: text
 ### Autobatching Enabled = true
 ### Processing batch Some(0):
 [0,]
-{uid: 0, details: {"receivedDocuments":2,"indexedDocuments":null}, stats: {"totalNbTasks":2,"status":{"enqueued":2},"types":{"documentAdditionOrUpdate":2},"indexUids":{"doggos":2}}, }
+{uid: 0, details: {"receivedDocuments":1,"indexedDocuments":null}, stats: {"totalNbTasks":1,"status":{"processing":1},"types":{"documentAdditionOrUpdate":1},"indexUids":{"doggos":1}}, }
 ----------------------------------------------------------------------
 ### All Tasks:
 0 {uid: 0, status: enqueued, details: { received_documents: 1, indexed_documents: None }, kind: DocumentAdditionOrUpdate { index_uid: "doggos", primary_key: Some("id"), method: ReplaceDocuments, content_file: 00000000-0000-0000-0000-000000000000, documents_count: 1, allow_index_creation: true }}

crates/index-scheduler/src/snapshots/lib.rs/fail_in_process_batch_for_document_addition/document_addition_batch_created.snap

@@ -5,7 +5,7 @@ snapshot_kind: text
 ### Autobatching Enabled = true
 ### Processing batch Some(0):
 [0,]
-{uid: 0, details: {"receivedDocuments":2,"indexedDocuments":null}, stats: {"totalNbTasks":2,"status":{"enqueued":2},"types":{"documentAdditionOrUpdate":2},"indexUids":{"doggos":2}}, }
+{uid: 0, details: {"receivedDocuments":1,"indexedDocuments":null}, stats: {"totalNbTasks":1,"status":{"processing":1},"types":{"documentAdditionOrUpdate":1},"indexUids":{"doggos":1}}, }
 ----------------------------------------------------------------------
 ### All Tasks:
 0 {uid: 0, status: enqueued, details: { received_documents: 1, indexed_documents: None }, kind: DocumentAdditionOrUpdate { index_uid: "doggos", primary_key: Some("id"), method: ReplaceDocuments, content_file: 00000000-0000-0000-0000-000000000000, documents_count: 1, allow_index_creation: true }}

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

@@ -9,8 +9,8 @@ source: crates/index-scheduler/src/lib.rs
 0 {uid: 0, batch_uid: 0, status: succeeded, details: { settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: Set({"catto"}), sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: NotSet, search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> } }, kind: SettingsUpdate { index_uid: "doggos", new_settings: Settings { displayed_attributes: WildcardSetting(NotSet), searchable_attributes: WildcardSetting(NotSet), filterable_attributes: Set({"catto"}), sortable_attributes: NotSet, ranking_rules: NotSet, stop_words: NotSet, non_separator_tokens: NotSet, separator_tokens: NotSet, dictionary: NotSet, synonyms: NotSet, distinct_attribute: NotSet, proximity_precision: NotSet, typo_tolerance: NotSet, faceting: NotSet, pagination: NotSet, embedders: NotSet, search_cutoff_ms: NotSet, localized_attributes: NotSet, facet_search: NotSet, prefix_search: NotSet, _kind: PhantomData<meilisearch_types::settings::Unchecked> }, is_deletion: false, allow_index_creation: true }}
 1 {uid: 1, batch_uid: 1, status: succeeded, details: { received_documents: 3, indexed_documents: Some(3) }, kind: DocumentAdditionOrUpdate { index_uid: "doggos", primary_key: Some("id"), method: ReplaceDocuments, content_file: 00000000-0000-0000-0000-000000000000, documents_count: 3, allow_index_creation: true }}
 2 {uid: 2, batch_uid: 2, status: succeeded, details: { received_document_ids: 1, deleted_documents: Some(1) }, kind: DocumentDeletion { index_uid: "doggos", documents_ids: ["1"] }}
-3 {uid: 3, batch_uid: 2, status: failed, error: ResponseError { code: 200, message: "Invalid type for filter subexpression: expected: String, Array, found: true.", error_code: "invalid_document_filter", error_type: "invalid_request", error_link: "https://docs.meilisearch.com/errors#invalid_document_filter" }, details: { original_filter: true, deleted_documents: Some(0) }, kind: DocumentDeletionByFilter { index_uid: "doggos", filter_expr: Bool(true) }}
-4 {uid: 4, batch_uid: 2, status: failed, error: ResponseError { code: 200, message: "Attribute `id` is not filterable. Available filterable attributes are: `catto`.\n1:3 id = 2", error_code: "invalid_document_filter", error_type: "invalid_request", error_link: "https://docs.meilisearch.com/errors#invalid_document_filter" }, details: { original_filter: "id = 2", deleted_documents: Some(0) }, kind: DocumentDeletionByFilter { index_uid: "doggos", filter_expr: String("id = 2") }}
+3 {uid: 3, batch_uid: 2, status: failed, error: ResponseError { code: 200, message: "Index `doggos`: Invalid type for filter subexpression: expected: String, Array, found: true.", error_code: "invalid_document_filter", error_type: "invalid_request", error_link: "https://docs.meilisearch.com/errors#invalid_document_filter" }, details: { original_filter: true, deleted_documents: Some(0) }, kind: DocumentDeletionByFilter { index_uid: "doggos", filter_expr: Bool(true) }}
+4 {uid: 4, batch_uid: 2, status: failed, error: ResponseError { code: 200, message: "Index `doggos`: Attribute `id` is not filterable. Available filterable attributes are: `catto`.\n1:3 id = 2", error_code: "invalid_document_filter", error_type: "invalid_request", error_link: "https://docs.meilisearch.com/errors#invalid_document_filter" }, details: { original_filter: "id = 2", deleted_documents: Some(0) }, kind: DocumentDeletionByFilter { index_uid: "doggos", filter_expr: String("id = 2") }}
 5 {uid: 5, batch_uid: 2, status: succeeded, details: { original_filter: "catto EXISTS", deleted_documents: Some(1) }, kind: DocumentDeletionByFilter { index_uid: "doggos", filter_expr: String("catto EXISTS") }}
 ----------------------------------------------------------------------
 ### Status:

crates/index-scheduler/src/snapshots/lib.rs/fail_in_update_task_after_process_batch_success_for_document_addition/after_batch_succeeded.snap

@@ -5,7 +5,7 @@ snapshot_kind: text
 ### Autobatching Enabled = true
 ### Processing batch Some(0):
 [0,]
-{uid: 0, details: {"receivedDocuments":2,"indexedDocuments":null}, stats: {"totalNbTasks":2,"status":{"enqueued":2},"types":{"documentAdditionOrUpdate":2},"indexUids":{"doggos":2}}, }
+{uid: 0, details: {"receivedDocuments":1,"indexedDocuments":null}, stats: {"totalNbTasks":1,"status":{"processing":1},"types":{"documentAdditionOrUpdate":1},"indexUids":{"doggos":1}}, }
 ----------------------------------------------------------------------
 ### All Tasks:
 0 {uid: 0, status: enqueued, details: { received_documents: 1, indexed_documents: None }, kind: DocumentAdditionOrUpdate { index_uid: "doggos", primary_key: Some("id"), method: ReplaceDocuments, content_file: 00000000-0000-0000-0000-000000000000, documents_count: 1, allow_index_creation: true }}

crates/index-scheduler/src/snapshots/lib.rs/fail_in_update_task_after_process_batch_success_for_document_addition/after_failing_to_commit.snap

@@ -5,7 +5,7 @@ snapshot_kind: text
 ### Autobatching Enabled = true
 ### Processing batch Some(0):
 [0,]
-{uid: 0, details: {"receivedDocuments":2,"indexedDocuments":null}, stats: {"totalNbTasks":2,"status":{"enqueued":2},"types":{"documentAdditionOrUpdate":2},"indexUids":{"doggos":2}}, }
+{uid: 0, details: {"receivedDocuments":1,"indexedDocuments":null}, stats: {"totalNbTasks":1,"status":{"processing":1},"types":{"documentAdditionOrUpdate":1},"indexUids":{"doggos":1}}, }
 ----------------------------------------------------------------------
 ### All Tasks:
 0 {uid: 0, status: enqueued, details: { received_documents: 1, indexed_documents: None }, kind: DocumentAdditionOrUpdate { index_uid: "doggos", primary_key: Some("id"), method: ReplaceDocuments, content_file: 00000000-0000-0000-0000-000000000000, documents_count: 1, allow_index_creation: true }}

crates/index-scheduler/src/snapshots/lib.rs/insert_task_while_another_task_is_processing/after_batch_creation.snap

@@ -5,7 +5,7 @@ snapshot_kind: text
 ### Autobatching Enabled = true
 ### Processing batch Some(0):
 [0,]
-{uid: 0, details: {"primaryKey":"id"}, stats: {"totalNbTasks":2,"status":{"enqueued":2},"types":{"indexCreation":2},"indexUids":{"index_a":2}}, }
+{uid: 0, details: {"primaryKey":"id"}, stats: {"totalNbTasks":1,"status":{"processing":1},"types":{"indexCreation":1},"indexUids":{"index_a":1}}, }
 ----------------------------------------------------------------------
 ### All Tasks:
 0 {uid: 0, status: enqueued, details: { primary_key: Some("id") }, kind: IndexCreation { index_uid: "index_a", primary_key: Some("id") }}

crates/index-scheduler/src/snapshots/lib.rs/insert_task_while_another_task_is_processing/registered_the_second_task.snap

@@ -5,7 +5,7 @@ snapshot_kind: text
 ### Autobatching Enabled = true
 ### Processing batch Some(0):
 [0,]
-{uid: 0, details: {"primaryKey":"id"}, stats: {"totalNbTasks":2,"status":{"enqueued":2},"types":{"indexCreation":2},"indexUids":{"index_a":2}}, }
+{uid: 0, details: {"primaryKey":"id"}, stats: {"totalNbTasks":1,"status":{"processing":1},"types":{"indexCreation":1},"indexUids":{"index_a":1}}, }
 ----------------------------------------------------------------------
 ### All Tasks:
 0 {uid: 0, status: enqueued, details: { primary_key: Some("id") }, kind: IndexCreation { index_uid: "index_a", primary_key: Some("id") }}

crates/index-scheduler/src/snapshots/lib.rs/insert_task_while_another_task_is_processing/registered_the_third_task.snap

@@ -5,7 +5,7 @@ snapshot_kind: text
 ### Autobatching Enabled = true
 ### Processing batch Some(0):
 [0,]
-{uid: 0, details: {"primaryKey":"id"}, stats: {"totalNbTasks":2,"status":{"enqueued":2},"types":{"indexCreation":2},"indexUids":{"index_a":2}}, }
+{uid: 0, details: {"primaryKey":"id"}, stats: {"totalNbTasks":1,"status":{"processing":1},"types":{"indexCreation":1},"indexUids":{"index_a":1}}, }
 ----------------------------------------------------------------------
 ### All Tasks:
 0 {uid: 0, status: enqueued, details: { primary_key: Some("id") }, kind: IndexCreation { index_uid: "index_a", primary_key: Some("id") }}

crates/index-scheduler/src/snapshots/lib.rs/query_batches_simple/after-advancing-a-bit.snap

@@ -5,7 +5,7 @@ snapshot_kind: text
 ### Autobatching Enabled = true
 ### Processing batch Some(1):
 [1,]
-{uid: 1, details: {"primaryKey":"sheep"}, stats: {"totalNbTasks":2,"status":{"enqueued":2},"types":{"indexCreation":2},"indexUids":{"doggo":2}}, }
+{uid: 1, details: {"primaryKey":"sheep"}, stats: {"totalNbTasks":1,"status":{"processing":1},"types":{"indexCreation":1},"indexUids":{"doggo":1}}, }
 ----------------------------------------------------------------------
 ### All Tasks:
 0 {uid: 0, batch_uid: 0, status: succeeded, details: { primary_key: Some("mouse") }, kind: IndexCreation { index_uid: "catto", primary_key: Some("mouse") }}

crates/index-scheduler/src/snapshots/lib.rs/task_deletion_undeleteable/task_deletion_processing.snap

@@ -5,7 +5,7 @@ snapshot_kind: text
 ### Autobatching Enabled = true
 ### Processing batch Some(0):
 [3,]
-{uid: 0, details: {"matchedTasks":2,"deletedTasks":null,"originalFilter":"test_query"}, stats: {"totalNbTasks":1,"status":{"enqueued":1},"types":{"taskDeletion":1},"indexUids":{}}, }
+{uid: 0, details: {"matchedTasks":2,"deletedTasks":null,"originalFilter":"test_query"}, stats: {"totalNbTasks":1,"status":{"processing":1},"types":{"taskDeletion":1},"indexUids":{}}, }
 ----------------------------------------------------------------------
 ### All Tasks:
 0 {uid: 0, status: enqueued, details: { primary_key: Some("mouse") }, kind: IndexCreation { index_uid: "catto", primary_key: Some("mouse") }}

crates/index-scheduler/src/utils.rs

@@ -67,7 +67,7 @@ impl ProcessingBatch {
             task.batch_uid = Some(self.uid);
             // We don't store the statuses in the map since they're all enqueued but we must
             // still store them in the stats since that can be displayed.
-            *self.stats.status.entry(task.status).or_default() += 1;
+            *self.stats.status.entry(Status::Processing).or_default() += 1;
 
             self.kinds.insert(task.kind.as_kind());
             *self.stats.types.entry(task.kind.as_kind()).or_default() += 1;

crates/meilisearch-types/src/error.rs

@@ -279,6 +279,7 @@ InvalidSearchPage                     , InvalidRequest       , BAD_REQUEST ;
 InvalidSearchQ                        , InvalidRequest       , BAD_REQUEST ;
 InvalidFacetSearchQuery               , InvalidRequest       , BAD_REQUEST ;
 InvalidFacetSearchName                , InvalidRequest       , BAD_REQUEST ;
+FacetSearchDisabled                   , InvalidRequest       , BAD_REQUEST ;
 InvalidSearchVector                   , InvalidRequest       , BAD_REQUEST ;
 InvalidSearchShowMatchesPosition      , InvalidRequest       , BAD_REQUEST ;
 InvalidSearchShowRankingScore         , InvalidRequest       , BAD_REQUEST ;

crates/meilisearch/src/error.rs

@@ -4,6 +4,7 @@ use byte_unit::{Byte, UnitType};
 use meilisearch_types::document_formats::{DocumentFormatError, PayloadType};
 use meilisearch_types::error::{Code, ErrorCode, ResponseError};
 use meilisearch_types::index_uid::{IndexUid, IndexUidFormatError};
+use meilisearch_types::milli;
 use meilisearch_types::milli::OrderBy;
 use serde_json::Value;
 use tokio::task::JoinError;
@@ -62,8 +63,11 @@ pub enum MeilisearchHttpError {
     HeedError(#[from] meilisearch_types::heed::Error),
     #[error(transparent)]
     IndexScheduler(#[from] index_scheduler::Error),
-    #[error(transparent)]
-    Milli(#[from] meilisearch_types::milli::Error),
+    #[error("{}", match .index_name {
+        Some(name) if !name.is_empty() => format!("Index `{}`: {error}", name),
+        _ => format!("{error}")
+    })]
+    Milli { error: milli::Error, index_name: Option<String> },
     #[error(transparent)]
     Payload(#[from] PayloadError),
     #[error(transparent)]
@@ -76,6 +80,12 @@ pub enum MeilisearchHttpError {
     MissingSearchHybrid,
 }
 
+impl MeilisearchHttpError {
+    pub(crate) fn from_milli(error: milli::Error, index_name: Option<String>) -> Self {
+        Self::Milli { error, index_name }
+    }
+}
+
 impl ErrorCode for MeilisearchHttpError {
     fn error_code(&self) -> Code {
         match self {
@@ -95,7 +105,7 @@ impl ErrorCode for MeilisearchHttpError {
             MeilisearchHttpError::SerdeJson(_) => Code::Internal,
             MeilisearchHttpError::HeedError(_) => Code::Internal,
             MeilisearchHttpError::IndexScheduler(e) => e.error_code(),
-            MeilisearchHttpError::Milli(e) => e.error_code(),
+            MeilisearchHttpError::Milli { error, .. } => error.error_code(),
             MeilisearchHttpError::Payload(e) => e.error_code(),
             MeilisearchHttpError::FileStore(_) => Code::Internal,
             MeilisearchHttpError::DocumentFormat(e) => e.error_code(),

crates/meilisearch/src/lib.rs

@@ -395,6 +395,7 @@ fn import_dump(
     for index_reader in dump_reader.indexes()? {
         let mut index_reader = index_reader?;
         let metadata = index_reader.metadata();
+        let uid = metadata.uid.clone();
         tracing::info!("Importing index `{}`.", metadata.uid);
 
         let date = Some((metadata.created_at, metadata.updated_at));
@@ -432,7 +433,7 @@ fn import_dump(
         let reader = DocumentsBatchReader::from_reader(reader)?;
 
         let embedder_configs = index.embedding_configs(&wtxn)?;
-        let embedders = index_scheduler.embedders(embedder_configs)?;
+        let embedders = index_scheduler.embedders(uid, embedder_configs)?;
 
         let builder = milli::update::IndexDocuments::new(
             &mut wtxn,

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

@@ -185,7 +185,8 @@ pub async fn search(
 
     let index = index_scheduler.index(&index_uid)?;
     let features = index_scheduler.features();
-    let search_kind = search_kind(&search_query, &index_scheduler, &index, features)?;
+    let search_kind =
+        search_kind(&search_query, &index_scheduler, index_uid.to_string(), &index, features)?;
     let permit = search_queue.try_get_search_permit().await?;
     let search_result = tokio::task::spawn_blocking(move || {
         perform_facet_search(

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

@@ -5,7 +5,7 @@ use actix_web::web::Data;
 use actix_web::{web, HttpRequest, HttpResponse};
 use deserr::actix_web::{AwebJson, AwebQueryParameter};
 use deserr::{DeserializeError, Deserr, ValuePointerRef};
-use index_scheduler::IndexScheduler;
+use index_scheduler::{Error, IndexScheduler};
 use meilisearch_types::deserr::query_params::Param;
 use meilisearch_types::deserr::{immutable_field_error, DeserrJsonError, DeserrQueryParamError};
 use meilisearch_types::error::deserr_codes::*;
@@ -107,7 +107,10 @@ pub async fn list_indexes(
             if !filters.is_index_authorized(uid) {
                 return Ok(None);
             }
-            Ok(Some(IndexView::new(uid.to_string(), index)?))
+            Ok(Some(
+                IndexView::new(uid.to_string(), index)
+                    .map_err(|e| Error::from_milli(e, Some(uid.to_string())))?,
+            ))
         })?;
     // Won't cause to open all indexes because IndexView doesn't keep the `Index` opened.
     let indexes: Vec<IndexView> = indexes.into_iter().flatten().collect();

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

@@ -243,11 +243,19 @@ pub async fn search_with_url_query(
     let index = index_scheduler.index(&index_uid)?;
     let features = index_scheduler.features();
 
-    let search_kind = search_kind(&query, index_scheduler.get_ref(), &index, features)?;
+    let search_kind =
+        search_kind(&query, index_scheduler.get_ref(), index_uid.to_string(), &index, features)?;
     let retrieve_vector = RetrieveVectors::new(query.retrieve_vectors, features)?;
     let permit = search_queue.try_get_search_permit().await?;
     let search_result = tokio::task::spawn_blocking(move || {
-        perform_search(&index, query, search_kind, retrieve_vector, index_scheduler.features())
+        perform_search(
+            index_uid.to_string(),
+            &index,
+            query,
+            search_kind,
+            retrieve_vector,
+            index_scheduler.features(),
+        )
     })
     .await;
     permit.drop().await;
@@ -287,12 +295,20 @@ pub async fn search_with_post(
 
     let features = index_scheduler.features();
 
-    let search_kind = search_kind(&query, index_scheduler.get_ref(), &index, features)?;
+    let search_kind =
+        search_kind(&query, index_scheduler.get_ref(), index_uid.to_string(), &index, features)?;
     let retrieve_vectors = RetrieveVectors::new(query.retrieve_vectors, features)?;
 
     let permit = search_queue.try_get_search_permit().await?;
     let search_result = tokio::task::spawn_blocking(move || {
-        perform_search(&index, query, search_kind, retrieve_vectors, index_scheduler.features())
+        perform_search(
+            index_uid.to_string(),
+            &index,
+            query,
+            search_kind,
+            retrieve_vectors,
+            index_scheduler.features(),
+        )
     })
     .await;
     permit.drop().await;
@@ -314,6 +330,7 @@ pub async fn search_with_post(
 pub fn search_kind(
     query: &SearchQuery,
     index_scheduler: &IndexScheduler,
+    index_uid: String,
     index: &milli::Index,
     features: RoFeatures,
 ) -> Result<SearchKind, ResponseError> {
@@ -332,7 +349,7 @@ pub fn search_kind(
         (None, _, None) => Ok(SearchKind::KeywordOnly),
         // hybrid.semantic_ratio == 1.0 => vector
         (_, Some(HybridQuery { semantic_ratio, embedder }), v) if **semantic_ratio == 1.0 => {
-            SearchKind::semantic(index_scheduler, index, embedder, v.map(|v| v.len()))
+            SearchKind::semantic(index_scheduler, index_uid, index, embedder, v.map(|v| v.len()))
         }
         // hybrid.semantic_ratio == 0.0 => keyword
         (_, Some(HybridQuery { semantic_ratio, embedder: _ }), _) if **semantic_ratio == 0.0 => {
@@ -340,13 +357,14 @@ pub fn search_kind(
         }
         // no query, hybrid, vector => semantic
         (None, Some(HybridQuery { semantic_ratio: _, embedder }), Some(v)) => {
-            SearchKind::semantic(index_scheduler, index, embedder, Some(v.len()))
+            SearchKind::semantic(index_scheduler, index_uid, index, embedder, Some(v.len()))
         }
         // query, no hybrid, no vector => keyword
         (Some(_), None, None) => Ok(SearchKind::KeywordOnly),
         // query, hybrid, maybe vector => hybrid
         (Some(_), Some(HybridQuery { semantic_ratio, embedder }), v) => SearchKind::hybrid(
             index_scheduler,
+            index_uid,
             index,
             embedder,
             **semantic_ratio,

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

@@ -103,8 +103,13 @@ async fn similar(
 
     let index = index_scheduler.index(&index_uid)?;
 
-    let (embedder_name, embedder, quantized) =
-        SearchKind::embedder(&index_scheduler, &index, &query.embedder, None)?;
+    let (embedder_name, embedder, quantized) = SearchKind::embedder(
+        &index_scheduler,
+        index_uid.to_string(),
+        &index,
+        &query.embedder,
+        None,
+    )?;
 
     tokio::task::spawn_blocking(move || {
         perform_similar(

crates/meilisearch/src/routes/multi_search.rs

@@ -125,14 +125,28 @@ pub async fn multi_search_with_post(
                         })
                         .with_index(query_index)?;
 
-                    let search_kind =
-                        search_kind(&query, index_scheduler.get_ref(), &index, features)
-                            .with_index(query_index)?;
+                    let index_uid_str = index_uid.to_string();
+
+                    let search_kind = search_kind(
+                        &query,
+                        index_scheduler.get_ref(),
+                        index_uid_str.clone(),
+                        &index,
+                        features,
+                    )
+                    .with_index(query_index)?;
                     let retrieve_vector = RetrieveVectors::new(query.retrieve_vectors, features)
                         .with_index(query_index)?;
 
                     let search_result = tokio::task::spawn_blocking(move || {
-                        perform_search(&index, query, search_kind, retrieve_vector, features)
+                        perform_search(
+                            index_uid_str.clone(),
+                            &index,
+                            query,
+                            search_kind,
+                            retrieve_vector,
+                            features,
+                        )
                     })
                     .await
                     .with_index(query_index)?;

crates/meilisearch/src/search/federated.rs

@@ -560,7 +560,8 @@ pub fn perform_federated_search(
             // use an immediately invoked lambda to capture the result without returning from the function
 
             let res: Result<(), ResponseError> = (|| {
-                let search_kind = search_kind(&query, index_scheduler, &index, features)?;
+                let search_kind =
+                    search_kind(&query, index_scheduler, index_uid.to_string(), &index, features)?;
 
                 let canonicalization_kind = match (&search_kind, &query.q) {
                     (SearchKind::SemanticOnly { .. }, _) => {
@@ -636,7 +637,8 @@ pub fn perform_federated_search(
                 search.offset(0);
                 search.limit(required_hit_count);
 
-                let (result, _semantic_hit_count) = super::search_from_kind(search_kind, search)?;
+                let (result, _semantic_hit_count) =
+                    super::search_from_kind(index_uid.to_string(), search_kind, search)?;
                 let format = AttributesFormat {
                     attributes_to_retrieve: query.attributes_to_retrieve,
                     retrieve_vectors,
@@ -670,7 +672,10 @@ pub fn perform_federated_search(
 
                 let formatter_builder = HitMaker::formatter_builder(matching_words, tokenizer);
 
-                let hit_maker = HitMaker::new(&index, &rtxn, format, formatter_builder)?;
+                let hit_maker =
+                    HitMaker::new(&index, &rtxn, format, formatter_builder).map_err(|e| {
+                        MeilisearchHttpError::from_milli(e, Some(index_uid.to_string()))
+                    })?;
 
                 results_by_query.push(SearchResultByQuery {
                     federation_options,

crates/meilisearch/src/search/mod.rs

@@ -19,7 +19,9 @@ use meilisearch_types::locales::Locale;
 use meilisearch_types::milli::score_details::{ScoreDetails, ScoringStrategy};
 use meilisearch_types::milli::vector::parsed_vectors::ExplicitVectors;
 use meilisearch_types::milli::vector::Embedder;
-use meilisearch_types::milli::{FacetValueHit, OrderBy, SearchForFacetValues, TimeBudget};
+use meilisearch_types::milli::{
+    FacetValueHit, InternalError, OrderBy, SearchForFacetValues, TimeBudget,
+};
 use meilisearch_types::settings::DEFAULT_PAGINATION_MAX_TOTAL_HITS;
 use meilisearch_types::{milli, Document};
 use milli::tokenizer::{Language, TokenizerBuilder};
@@ -281,35 +283,38 @@ pub enum SearchKind {
 impl SearchKind {
     pub(crate) fn semantic(
         index_scheduler: &index_scheduler::IndexScheduler,
+        index_uid: String,
         index: &Index,
         embedder_name: &str,
         vector_len: Option<usize>,
     ) -> Result<Self, ResponseError> {
         let (embedder_name, embedder, quantized) =
-            Self::embedder(index_scheduler, index, embedder_name, vector_len)?;
+            Self::embedder(index_scheduler, index_uid, index, embedder_name, vector_len)?;
         Ok(Self::SemanticOnly { embedder_name, embedder, quantized })
     }
 
     pub(crate) fn hybrid(
         index_scheduler: &index_scheduler::IndexScheduler,
+        index_uid: String,
         index: &Index,
         embedder_name: &str,
         semantic_ratio: f32,
         vector_len: Option<usize>,
     ) -> Result<Self, ResponseError> {
         let (embedder_name, embedder, quantized) =
-            Self::embedder(index_scheduler, index, embedder_name, vector_len)?;
+            Self::embedder(index_scheduler, index_uid, index, embedder_name, vector_len)?;
         Ok(Self::Hybrid { embedder_name, embedder, quantized, semantic_ratio })
     }
 
     pub(crate) fn embedder(
         index_scheduler: &index_scheduler::IndexScheduler,
+        index_uid: String,
         index: &Index,
         embedder_name: &str,
         vector_len: Option<usize>,
     ) -> Result<(String, Arc<Embedder>, bool), ResponseError> {
         let embedder_configs = index.embedding_configs(&index.read_txn()?)?;
-        let embedders = index_scheduler.embedders(embedder_configs)?;
+        let embedders = index_scheduler.embedders(index_uid, embedder_configs)?;
 
         let (embedder, _, quantized) = embedders
             .get(embedder_name)
@@ -890,6 +895,7 @@ fn prepare_search<'t>(
 }
 
 pub fn perform_search(
+    index_uid: String,
     index: &Index,
     query: SearchQuery,
     search_kind: SearchKind,
@@ -916,7 +922,7 @@ pub fn perform_search(
             used_negative_operator,
         },
         semantic_hit_count,
-    ) = search_from_kind(search_kind, search)?;
+    ) = search_from_kind(index_uid, search_kind, search)?;
 
     let SearchQuery {
         q,
@@ -1069,17 +1075,27 @@ fn compute_facet_distribution_stats<S: AsRef<str>>(
 }
 
 pub fn search_from_kind(
+    index_uid: String,
     search_kind: SearchKind,
     search: milli::Search<'_>,
 ) -> Result<(milli::SearchResult, Option<u32>), MeilisearchHttpError> {
     let (milli_result, semantic_hit_count) = match &search_kind {
-        SearchKind::KeywordOnly => (search.execute()?, None),
+        SearchKind::KeywordOnly => {
+            let results = search
+                .execute()
+                .map_err(|e| MeilisearchHttpError::from_milli(e, Some(index_uid.to_string())))?;
+            (results, None)
+        }
         SearchKind::SemanticOnly { .. } => {
-            let results = search.execute()?;
+            let results = search
+                .execute()
+                .map_err(|e| MeilisearchHttpError::from_milli(e, Some(index_uid.to_string())))?;
             let semantic_hit_count = results.document_scores.len() as u32;
             (results, Some(semantic_hit_count))
         }
-        SearchKind::Hybrid { semantic_ratio, .. } => search.execute_hybrid(*semantic_ratio)?,
+        SearchKind::Hybrid { semantic_ratio, .. } => search
+            .execute_hybrid(*semantic_ratio)
+            .map_err(|e| MeilisearchHttpError::from_milli(e, Some(index_uid)))?,
     };
     Ok((milli_result, semantic_hit_count))
 }
@@ -1181,7 +1197,7 @@ impl<'a> HitMaker<'a> {
         rtxn: &'a RoTxn<'a>,
         format: AttributesFormat,
         mut formatter_builder: MatcherBuilder<'a>,
-    ) -> Result<Self, MeilisearchHttpError> {
+    ) -> milli::Result<Self> {
         formatter_builder.crop_marker(format.crop_marker);
         formatter_builder.highlight_prefix(format.highlight_pre_tag);
         formatter_builder.highlight_suffix(format.highlight_post_tag);
@@ -1276,11 +1292,7 @@ impl<'a> HitMaker<'a> {
         })
     }
 
-    pub fn make_hit(
-        &self,
-        id: u32,
-        score: &[ScoreDetails],
-    ) -> Result<SearchHit, MeilisearchHttpError> {
+    pub fn make_hit(&self, id: u32, score: &[ScoreDetails]) -> milli::Result<SearchHit> {
         let (_, obkv) =
             self.index.iter_documents(self.rtxn, std::iter::once(id))?.next().unwrap()?;
 
@@ -1323,7 +1335,10 @@ impl<'a> HitMaker<'a> {
                     .is_some_and(|conf| conf.user_provided.contains(id));
                 let embeddings =
                     ExplicitVectors { embeddings: Some(vector.into()), regenerate: !user_provided };
-                vectors.insert(name, serde_json::to_value(embeddings)?);
+                vectors.insert(
+                    name,
+                    serde_json::to_value(embeddings).map_err(InternalError::SerdeJson)?,
+                );
             }
             document.insert("_vectors".into(), vectors.into());
         }
@@ -1369,7 +1384,7 @@ fn make_hits<'a>(
     format: AttributesFormat,
     matching_words: milli::MatchingWords,
     documents_ids_scores: impl Iterator<Item = (u32, &'a Vec<ScoreDetails>)> + 'a,
-) -> Result<Vec<SearchHit>, MeilisearchHttpError> {
+) -> milli::Result<Vec<SearchHit>> {
     let mut documents = Vec::new();
 
     let dictionary = index.dictionary(rtxn)?;
@@ -1407,6 +1422,13 @@ pub fn perform_facet_search(
         None => TimeBudget::default(),
     };
 
+    if !index.facet_search(&rtxn)? {
+        return Err(ResponseError::from_msg(
+            "The facet search is disabled for this index".to_string(),
+            Code::FacetSearchDisabled,
+        ));
+    }
+
     // In the faceted search context, we want to use the intersection between the locales provided by the user
     // and the locales of the facet string.
     // If the facet string is not localized, we **ignore** the locales provided by the user because the facet data has no locale.
@@ -1690,12 +1712,12 @@ fn make_document(
     displayed_attributes: &BTreeSet<FieldId>,
     field_ids_map: &FieldsIdsMap,
     obkv: &obkv::KvReaderU16,
-) -> Result<Document, MeilisearchHttpError> {
+) -> milli::Result<Document> {
     let mut document = serde_json::Map::new();
 
     // recreate the original json
     for (key, value) in obkv.iter() {
-        let value = serde_json::from_slice(value)?;
+        let value = serde_json::from_slice(value).map_err(InternalError::SerdeJson)?;
         let key = field_ids_map.name(key).expect("Missing field name").to_string();
 
         document.insert(key, value);
@@ -1720,7 +1742,7 @@ fn format_fields(
     displayable_ids: &BTreeSet<FieldId>,
     locales: Option<&[Language]>,
     localized_attributes: &[LocalizedAttributesRule],
-) -> Result<(Option<MatchesPosition>, Document), MeilisearchHttpError> {
+) -> milli::Result<(Option<MatchesPosition>, Document)> {
     let mut matches_position = compute_matches.then(BTreeMap::new);
     let mut document = document.clone();
 
@@ -1898,7 +1920,7 @@ fn parse_filter_array(arr: &[Value]) -> Result<Option<Filter>, MeilisearchHttpEr
         }
     }
 
-    Ok(Filter::from_array(ands)?)
+    Filter::from_array(ands).map_err(|e| MeilisearchHttpError::from_milli(e, None))
 }
 
 #[cfg(test)]

crates/meilisearch/tests/common/mod.rs

@@ -52,6 +52,25 @@ impl Value {
         }
         self
     }
+
+    /// Return `true` if the `status` field is set to `failed`.
+    /// Panic if the `status` field doesn't exists.
+    #[track_caller]
+    pub fn is_fail(&self) -> bool {
+        if !self["status"].is_string() {
+            panic!("Called `is_fail` on {}", serde_json::to_string_pretty(&self.0).unwrap());
+        }
+        self["status"] == serde_json::Value::String(String::from("failed"))
+    }
+
+    // Panic if the json doesn't contain the `status` field set to "succeeded"
+    #[track_caller]
+    pub fn failed(&self) -> &Self {
+        if !self.is_fail() {
+            panic!("Called failed on {}", serde_json::to_string_pretty(&self.0).unwrap());
+        }
+        self
+    }
 }
 
 impl From<serde_json::Value> for Value {

crates/meilisearch/tests/documents/add_documents.rs

@@ -1681,7 +1681,7 @@ async fn add_documents_invalid_geo_field() {
         "indexedDocuments": 0
       },
       "error": {
-        "message": "The `_geo` field in the document with the id: `\"11\"` is not an object. Was expecting an object with the `_geo.lat` and `_geo.lng` fields but instead got `\"foobar\"`.",
+        "message": "Index `test`: The `_geo` field in the document with the id: `\"11\"` is not an object. Was expecting an object with the `_geo.lat` and `_geo.lng` fields but instead got `\"foobar\"`.",
         "code": "invalid_document_geo_field",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#invalid_document_geo_field"
@@ -1719,7 +1719,7 @@ async fn add_documents_invalid_geo_field() {
         "indexedDocuments": 0
       },
       "error": {
-        "message": "Could not find latitude nor longitude in the document with the id: `\"11\"`. Was expecting `_geo.lat` and `_geo.lng` fields.",
+        "message": "Index `test`: Could not find latitude nor longitude in the document with the id: `\"11\"`. Was expecting `_geo.lat` and `_geo.lng` fields.",
         "code": "invalid_document_geo_field",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#invalid_document_geo_field"
@@ -1757,7 +1757,7 @@ async fn add_documents_invalid_geo_field() {
         "indexedDocuments": 0
       },
       "error": {
-        "message": "Could not find latitude nor longitude in the document with the id: `\"11\"`. Was expecting `_geo.lat` and `_geo.lng` fields.",
+        "message": "Index `test`: Could not find latitude nor longitude in the document with the id: `\"11\"`. Was expecting `_geo.lat` and `_geo.lng` fields.",
         "code": "invalid_document_geo_field",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#invalid_document_geo_field"
@@ -1795,7 +1795,7 @@ async fn add_documents_invalid_geo_field() {
         "indexedDocuments": 0
       },
       "error": {
-        "message": "Could not find longitude in the document with the id: `\"11\"`. Was expecting a `_geo.lng` field.",
+        "message": "Index `test`: Could not find longitude in the document with the id: `\"11\"`. Was expecting a `_geo.lng` field.",
         "code": "invalid_document_geo_field",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#invalid_document_geo_field"
@@ -1833,7 +1833,7 @@ async fn add_documents_invalid_geo_field() {
         "indexedDocuments": 0
       },
       "error": {
-        "message": "Could not find latitude in the document with the id: `\"11\"`. Was expecting a `_geo.lat` field.",
+        "message": "Index `test`: Could not find latitude in the document with the id: `\"11\"`. Was expecting a `_geo.lat` field.",
         "code": "invalid_document_geo_field",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#invalid_document_geo_field"
@@ -1871,7 +1871,7 @@ async fn add_documents_invalid_geo_field() {
         "indexedDocuments": 0
       },
       "error": {
-        "message": "Could not find longitude in the document with the id: `\"11\"`. Was expecting a `_geo.lng` field.",
+        "message": "Index `test`: Could not find longitude in the document with the id: `\"11\"`. Was expecting a `_geo.lng` field.",
         "code": "invalid_document_geo_field",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#invalid_document_geo_field"
@@ -1909,7 +1909,7 @@ async fn add_documents_invalid_geo_field() {
         "indexedDocuments": 0
       },
       "error": {
-        "message": "Could not find latitude in the document with the id: `\"11\"`. Was expecting a `_geo.lat` field.",
+        "message": "Index `test`: Could not find latitude in the document with the id: `\"11\"`. Was expecting a `_geo.lat` field.",
         "code": "invalid_document_geo_field",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#invalid_document_geo_field"
@@ -1947,7 +1947,7 @@ async fn add_documents_invalid_geo_field() {
         "indexedDocuments": 0
       },
       "error": {
-        "message": "Could not parse latitude nor longitude in the document with the id: `\"11\"`. Was expecting finite numbers but instead got `false` and `true`.",
+        "message": "Index `test`: Could not parse latitude nor longitude in the document with the id: `\"11\"`. Was expecting finite numbers but instead got `false` and `true`.",
         "code": "invalid_document_geo_field",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#invalid_document_geo_field"
@@ -1985,7 +1985,7 @@ async fn add_documents_invalid_geo_field() {
         "indexedDocuments": 0
       },
       "error": {
-        "message": "Could not find longitude in the document with the id: `\"11\"`. Was expecting a `_geo.lng` field.",
+        "message": "Index `test`: Could not find longitude in the document with the id: `\"11\"`. Was expecting a `_geo.lng` field.",
         "code": "invalid_document_geo_field",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#invalid_document_geo_field"
@@ -2023,7 +2023,7 @@ async fn add_documents_invalid_geo_field() {
         "indexedDocuments": 0
       },
       "error": {
-        "message": "Could not find latitude in the document with the id: `\"11\"`. Was expecting a `_geo.lat` field.",
+        "message": "Index `test`: Could not find latitude in the document with the id: `\"11\"`. Was expecting a `_geo.lat` field.",
         "code": "invalid_document_geo_field",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#invalid_document_geo_field"
@@ -2061,7 +2061,7 @@ async fn add_documents_invalid_geo_field() {
         "indexedDocuments": 0
       },
       "error": {
-        "message": "Could not parse latitude nor longitude in the document with the id: `\"11\"`. Was expecting finite numbers but instead got `\"doggo\"` and `\"doggo\"`.",
+        "message": "Index `test`: Could not parse latitude nor longitude in the document with the id: `\"11\"`. Was expecting finite numbers but instead got `\"doggo\"` and `\"doggo\"`.",
         "code": "invalid_document_geo_field",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#invalid_document_geo_field"
@@ -2099,7 +2099,7 @@ async fn add_documents_invalid_geo_field() {
         "indexedDocuments": 0
       },
       "error": {
-        "message": "The `_geo` field in the document with the id: `\"11\"` contains the following unexpected fields: `{\"doggo\":\"are the best\"}`.",
+        "message": "Index `test`: The `_geo` field in the document with the id: `\"11\"` contains the following unexpected fields: `{\"doggo\":\"are the best\"}`.",
         "code": "invalid_document_geo_field",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#invalid_document_geo_field"
@@ -2138,7 +2138,7 @@ async fn add_documents_invalid_geo_field() {
         "indexedDocuments": 0
       },
       "error": {
-        "message": "Could not parse longitude in the document with the id: `\"12\"`. Was expecting a finite number but instead got `null`.",
+        "message": "Index `test`: Could not parse longitude in the document with the id: `\"12\"`. Was expecting a finite number but instead got `null`.",
         "code": "invalid_document_geo_field",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#invalid_document_geo_field"
@@ -2175,7 +2175,7 @@ async fn add_documents_invalid_geo_field() {
         "indexedDocuments": 0
       },
       "error": {
-        "message": "Could not parse latitude in the document with the id: `\"12\"`. Was expecting a finite number but instead got `null`.",
+        "message": "Index `test`: Could not parse latitude in the document with the id: `\"12\"`. Was expecting a finite number but instead got `null`.",
         "code": "invalid_document_geo_field",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#invalid_document_geo_field"
@@ -2212,7 +2212,7 @@ async fn add_documents_invalid_geo_field() {
         "indexedDocuments": 0
       },
       "error": {
-        "message": "Could not parse latitude nor longitude in the document with the id: `\"13\"`. Was expecting finite numbers but instead got `null` and `null`.",
+        "message": "Index `test`: Could not parse latitude nor longitude in the document with the id: `\"13\"`. Was expecting finite numbers but instead got `null` and `null`.",
         "code": "invalid_document_geo_field",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#invalid_document_geo_field"
@@ -2279,7 +2279,7 @@ async fn add_invalid_geo_and_then_settings() {
         ]
       },
       "error": {
-        "message": "Could not parse latitude in the document with the id: `\"11\"`. Was expecting a finite number but instead got `null`.",
+        "message": "Index `test`: Could not parse latitude in the document with the id: `\"11\"`. Was expecting a finite number but instead got `null`.",
         "code": "invalid_document_geo_field",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#invalid_document_geo_field"

crates/meilisearch/tests/documents/errors.rs

@@ -604,7 +604,7 @@ async fn delete_document_by_filter() {
         "originalFilter": "\"doggo = bernese\""
       },
       "error": {
-        "message": "Attribute `doggo` is not filterable. This index does not have configured filterable attributes.\n1:6 doggo = bernese",
+        "message": "Index `EMPTY_INDEX`: Attribute `doggo` is not filterable. This index does not have configured filterable attributes.\n1:6 doggo = bernese",
         "code": "invalid_document_filter",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#invalid_document_filter"
@@ -636,7 +636,7 @@ async fn delete_document_by_filter() {
         "originalFilter": "\"catto = jorts\""
       },
       "error": {
-        "message": "Attribute `catto` is not filterable. Available filterable attributes are: `id`, `title`.\n1:6 catto = jorts",
+        "message": "Index `SHARED_DOCUMENTS`: Attribute `catto` is not filterable. Available filterable attributes are: `id`, `title`.\n1:6 catto = jorts",
         "code": "invalid_document_filter",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#invalid_document_filter"

crates/meilisearch/tests/index/update_index.rs

@@ -95,7 +95,7 @@ async fn error_update_existing_primary_key() {
     let response = index.wait_task(2).await;
 
     let expected_response = json!({
-        "message": "Index already has a primary key: `id`.",
+        "message": "Index `test`: Index already has a primary key: `id`.",
         "code": "index_primary_key_already_exists",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#index_primary_key_already_exists"

crates/meilisearch/tests/search/errors.rs

@@ -711,7 +711,7 @@ async fn filter_invalid_attribute_array() {
     index.wait_task(task.uid()).await;
 
     let expected_response = json!({
-        "message": "Attribute `many` is not filterable. Available filterable attributes are: `title`.\n1:5 many = Glass",
+        "message": format!("Index `{}`: Attribute `many` is not filterable. Available filterable attributes are: `title`.\n1:5 many = Glass", index.uid),
         "code": "invalid_search_filter",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#invalid_search_filter"
@@ -733,7 +733,7 @@ async fn filter_invalid_attribute_string() {
     index.wait_task(task.uid()).await;
 
     let expected_response = json!({
-        "message": "Attribute `many` is not filterable. Available filterable attributes are: `title`.\n1:5 many = Glass",
+        "message": format!("Index `{}`: Attribute `many` is not filterable. Available filterable attributes are: `title`.\n1:5 many = Glass", index.uid),
         "code": "invalid_search_filter",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#invalid_search_filter"
@@ -940,7 +940,7 @@ async fn sort_unsortable_attribute() {
     index.wait_task(response.uid()).await.succeeded();
 
     let expected_response = json!({
-        "message": "Attribute `title` is not sortable. Available sortable attributes are: `id`.",
+        "message": format!("Index `{}`: Attribute `title` is not sortable. Available sortable attributes are: `id`.", index.uid),
         "code": "invalid_search_sort",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#invalid_search_sort"
@@ -998,7 +998,7 @@ async fn sort_unset_ranking_rule() {
     index.wait_task(response.uid()).await.succeeded();
 
     let expected_response = json!({
-        "message": "You must specify where `sort` is listed in the rankingRules setting to use the sort parameter at search time.",
+        "message": format!("Index `{}`: You must specify where `sort` is listed in the rankingRules setting to use the sort parameter at search time.", index.uid),
         "code": "invalid_search_sort",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#invalid_search_sort"
@@ -1024,19 +1024,18 @@ async fn search_on_unknown_field() {
         index.update_settings_searchable_attributes(json!(["id", "title"])).await;
     index.wait_task(response.uid()).await.succeeded();
 
+    let expected_response = json!({
+        "message": format!("Index `{}`: Attribute `unknown` is not searchable. Available searchable attributes are: `id, title`.", index.uid),
+        "code": "invalid_search_attributes_to_search_on",
+        "type": "invalid_request",
+        "link": "https://docs.meilisearch.com/errors#invalid_search_attributes_to_search_on"
+    });
     index
         .search(
             json!({"q": "Captain Marvel", "attributesToSearchOn": ["unknown"]}),
             |response, code| {
-                snapshot!(code, @"400 Bad Request");
-                snapshot!(json_string!(response), @r###"
-                {
-                  "message": "Attribute `unknown` is not searchable. Available searchable attributes are: `id, title`.",
-                  "code": "invalid_search_attributes_to_search_on",
-                  "type": "invalid_request",
-                  "link": "https://docs.meilisearch.com/errors#invalid_search_attributes_to_search_on"
-                }
-                "###);
+                assert_eq!(response, expected_response);
+                assert_eq!(code, 400);
             },
         )
         .await;
@@ -1050,19 +1049,18 @@ async fn search_on_unknown_field_plus_joker() {
         index.update_settings_searchable_attributes(json!(["id", "title"])).await;
     index.wait_task(response.uid()).await.succeeded();
 
+    let expected_response = json!({
+        "message": format!("Index `{}`: Attribute `unknown` is not searchable. Available searchable attributes are: `id, title`.", index.uid),
+        "code": "invalid_search_attributes_to_search_on",
+        "type": "invalid_request",
+        "link": "https://docs.meilisearch.com/errors#invalid_search_attributes_to_search_on"
+    });
     index
         .search(
             json!({"q": "Captain Marvel", "attributesToSearchOn": ["*", "unknown"]}),
             |response, code| {
-                snapshot!(code, @"400 Bad Request");
-                snapshot!(json_string!(response), @r###"
-                {
-                  "message": "Attribute `unknown` is not searchable. Available searchable attributes are: `id, title`.",
-                  "code": "invalid_search_attributes_to_search_on",
-                  "type": "invalid_request",
-                  "link": "https://docs.meilisearch.com/errors#invalid_search_attributes_to_search_on"
-                }
-                "###);
+                assert_eq!(response, expected_response);
+                assert_eq!(code, 400);
             },
         )
         .await;
@@ -1071,15 +1069,8 @@ async fn search_on_unknown_field_plus_joker() {
         .search(
             json!({"q": "Captain Marvel", "attributesToSearchOn": ["unknown", "*"]}),
             |response, code| {
-                snapshot!(code, @"400 Bad Request");
-                snapshot!(json_string!(response), @r###"
-                {
-                  "message": "Attribute `unknown` is not searchable. Available searchable attributes are: `id, title`.",
-                  "code": "invalid_search_attributes_to_search_on",
-                  "type": "invalid_request",
-                  "link": "https://docs.meilisearch.com/errors#invalid_search_attributes_to_search_on"
-                }
-                "###);
+                assert_eq!(response, expected_response);
+                assert_eq!(code, 400);
             },
         )
         .await;
@@ -1092,47 +1083,44 @@ async fn distinct_at_search_time() {
     let (task, _) = index.create(None).await;
     index.wait_task(task.uid()).await.succeeded();
 
+    let expected_response = json!({
+        "message": format!("Index `{}`: Attribute `doggo.truc` is not filterable and thus, cannot be used as distinct attribute. This index does not have configured filterable attributes.", index.uid),
+        "code": "invalid_search_distinct",
+        "type": "invalid_request",
+        "link": "https://docs.meilisearch.com/errors#invalid_search_distinct"
+    });
     let (response, code) =
         index.search_post(json!({"page": 0, "hitsPerPage": 2, "distinct": "doggo.truc"})).await;
-    snapshot!(code, @"400 Bad Request");
-    snapshot!(response, @r###"
-    {
-      "message": "Attribute `doggo.truc` is not filterable and thus, cannot be used as distinct attribute. This index does not have configured filterable attributes.",
-      "code": "invalid_search_distinct",
-      "type": "invalid_request",
-      "link": "https://docs.meilisearch.com/errors#invalid_search_distinct"
-    }
-    "###);
+    assert_eq!(response, expected_response);
+    assert_eq!(code, 400);
 
     let (task, _) = index.update_settings_filterable_attributes(json!(["color", "machin"])).await;
     index.wait_task(task.uid()).await;
 
+    let expected_response = json!({
+        "message": format!("Index `{}`: Attribute `doggo.truc` is not filterable and thus, cannot be used as distinct attribute. Available filterable attributes are: `color, machin`.", index.uid),
+        "code": "invalid_search_distinct",
+        "type": "invalid_request",
+        "link": "https://docs.meilisearch.com/errors#invalid_search_distinct"
+    });
     let (response, code) =
         index.search_post(json!({"page": 0, "hitsPerPage": 2, "distinct": "doggo.truc"})).await;
-    snapshot!(code, @"400 Bad Request");
-    snapshot!(response, @r###"
-    {
-      "message": "Attribute `doggo.truc` is not filterable and thus, cannot be used as distinct attribute. Available filterable attributes are: `color, machin`.",
-      "code": "invalid_search_distinct",
-      "type": "invalid_request",
-      "link": "https://docs.meilisearch.com/errors#invalid_search_distinct"
-    }
-    "###);
+    assert_eq!(response, expected_response);
+    assert_eq!(code, 400);
 
     let (task, _) = index.update_settings_displayed_attributes(json!(["color"])).await;
     index.wait_task(task.uid()).await;
 
+    let expected_response = json!({
+        "message": format!("Index `{}`: Attribute `doggo.truc` is not filterable and thus, cannot be used as distinct attribute. Available filterable attributes are: `color, <..hidden-attributes>`.", index.uid),
+        "code": "invalid_search_distinct",
+        "type": "invalid_request",
+        "link": "https://docs.meilisearch.com/errors#invalid_search_distinct"
+    });
     let (response, code) =
         index.search_post(json!({"page": 0, "hitsPerPage": 2, "distinct": "doggo.truc"})).await;
-    snapshot!(code, @"400 Bad Request");
-    snapshot!(response, @r###"
-    {
-      "message": "Attribute `doggo.truc` is not filterable and thus, cannot be used as distinct attribute. Available filterable attributes are: `color, <..hidden-attributes>`.",
-      "code": "invalid_search_distinct",
-      "type": "invalid_request",
-      "link": "https://docs.meilisearch.com/errors#invalid_search_distinct"
-    }
-    "###);
+    assert_eq!(response, expected_response);
+    assert_eq!(code, 400);
 
     let (response, code) =
         index.search_post(json!({"page": 0, "hitsPerPage": 2, "distinct": true})).await;

crates/meilisearch/tests/search/facet_search.rs

@@ -221,8 +221,15 @@ async fn add_documents_and_deactivate_facet_search() {
     let (response, code) =
         index.facet_search(json!({"facetName": "genres", "facetQuery": "a"})).await;
 
-    assert_eq!(code, 200, "{}", response);
-    assert_eq!(dbg!(response)["facetHits"].as_array().unwrap().len(), 0);
+    assert_eq!(code, 400, "{}", response);
+    snapshot!(response, @r###"
+    {
+      "message": "The facet search is disabled for this index",
+      "code": "facet_search_disabled",
+      "type": "invalid_request",
+      "link": "https://docs.meilisearch.com/errors#facet_search_disabled"
+    }
+    "###);
 }
 
 #[actix_rt::test]
@@ -245,8 +252,15 @@ async fn deactivate_facet_search_and_add_documents() {
     let (response, code) =
         index.facet_search(json!({"facetName": "genres", "facetQuery": "a"})).await;
 
-    assert_eq!(code, 200, "{}", response);
-    assert_eq!(dbg!(response)["facetHits"].as_array().unwrap().len(), 0);
+    assert_eq!(code, 400, "{}", response);
+    snapshot!(response, @r###"
+    {
+      "message": "The facet search is disabled for this index",
+      "code": "facet_search_disabled",
+      "type": "invalid_request",
+      "link": "https://docs.meilisearch.com/errors#facet_search_disabled"
+    }
+    "###);
 }
 
 #[actix_rt::test]

crates/meilisearch/tests/search/multi.rs

@@ -1070,7 +1070,7 @@ async fn federation_one_query_error() {
     snapshot!(code, @"400 Bad Request");
     snapshot!(json_string!(response), @r###"
     {
-      "message": "Inside `.queries[1]`: Attribute `title` is not filterable. This index does not have configured filterable attributes.\n1:6 title = toto",
+      "message": "Inside `.queries[1]`: Index `nested`: Attribute `title` is not filterable. This index does not have configured filterable attributes.\n1:6 title = toto",
       "code": "invalid_search_filter",
       "type": "invalid_request",
       "link": "https://docs.meilisearch.com/errors#invalid_search_filter"
@@ -1102,7 +1102,7 @@ async fn federation_one_query_sort_error() {
     snapshot!(code, @"400 Bad Request");
     snapshot!(json_string!(response), @r###"
     {
-      "message": "Inside `.queries[1]`: Attribute `doggos` is not sortable. This index does not have configured sortable attributes.",
+      "message": "Inside `.queries[1]`: Index `nested`: Attribute `doggos` is not sortable. This index does not have configured sortable attributes.",
       "code": "invalid_search_sort",
       "type": "invalid_request",
       "link": "https://docs.meilisearch.com/errors#invalid_search_sort"
@@ -1166,7 +1166,7 @@ async fn federation_multiple_query_errors() {
     snapshot!(code, @"400 Bad Request");
     snapshot!(json_string!(response), @r###"
     {
-      "message": "Inside `.queries[0]`: Attribute `title` is not filterable. This index does not have configured filterable attributes.\n1:6 title = toto",
+      "message": "Inside `.queries[0]`: Index `test`: Attribute `title` is not filterable. This index does not have configured filterable attributes.\n1:6 title = toto",
       "code": "invalid_search_filter",
       "type": "invalid_request",
       "link": "https://docs.meilisearch.com/errors#invalid_search_filter"
@@ -1198,7 +1198,7 @@ async fn federation_multiple_query_sort_errors() {
     snapshot!(code, @"400 Bad Request");
     snapshot!(json_string!(response), @r###"
     {
-      "message": "Inside `.queries[0]`: Attribute `title` is not sortable. This index does not have configured sortable attributes.",
+      "message": "Inside `.queries[0]`: Index `test`: Attribute `title` is not sortable. This index does not have configured sortable attributes.",
       "code": "invalid_search_sort",
       "type": "invalid_request",
       "link": "https://docs.meilisearch.com/errors#invalid_search_sort"
@@ -1231,7 +1231,7 @@ async fn federation_multiple_query_errors_interleaved() {
     snapshot!(code, @"400 Bad Request");
     snapshot!(json_string!(response), @r###"
     {
-      "message": "Inside `.queries[1]`: Attribute `doggos` is not filterable. This index does not have configured filterable attributes.\n1:7 doggos IN [intel, kefir]",
+      "message": "Inside `.queries[1]`: Index `nested`: Attribute `doggos` is not filterable. This index does not have configured filterable attributes.\n1:7 doggos IN [intel, kefir]",
       "code": "invalid_search_filter",
       "type": "invalid_request",
       "link": "https://docs.meilisearch.com/errors#invalid_search_filter"
@@ -1264,7 +1264,7 @@ async fn federation_multiple_query_sort_errors_interleaved() {
     snapshot!(code, @"400 Bad Request");
     snapshot!(json_string!(response), @r###"
     {
-      "message": "Inside `.queries[1]`: Attribute `doggos` is not sortable. This index does not have configured sortable attributes.",
+      "message": "Inside `.queries[1]`: Index `nested`: Attribute `doggos` is not sortable. This index does not have configured sortable attributes.",
       "code": "invalid_search_sort",
       "type": "invalid_request",
       "link": "https://docs.meilisearch.com/errors#invalid_search_sort"

crates/meilisearch/tests/snapshot/mod.rs

@@ -129,11 +129,11 @@ async fn perform_on_demand_snapshot() {
 
     index.load_test_set().await;
 
-    server.index("doggo").create(Some("bone")).await;
-    index.wait_task(2).await;
+    let (task, _) = server.index("doggo").create(Some("bone")).await;
+    index.wait_task(task.uid()).await.succeeded();
 
-    server.index("doggo").create(Some("bone")).await;
-    index.wait_task(2).await;
+    let (task, _) = server.index("doggo").create(Some("bone")).await;
+    index.wait_task(task.uid()).await.failed();
 
     let (task, code) = server.create_snapshot().await;
     snapshot!(code, @"202 Accepted");

crates/meilisearch/tests/tasks/mod.rs

@@ -448,7 +448,7 @@ async fn test_summarized_delete_documents_by_filter() {
         "originalFilter": "\"doggo = bernese\""
       },
       "error": {
-        "message": "Attribute `doggo` is not filterable. This index does not have configured filterable attributes.\n1:6 doggo = bernese",
+        "message": "Index `test`: Attribute `doggo` is not filterable. This index does not have configured filterable attributes.\n1:6 doggo = bernese",
         "code": "invalid_document_filter",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#invalid_document_filter"

crates/meilisearch/tests/vector/binary_quantized.rs

@@ -318,7 +318,7 @@ async fn try_to_disable_binary_quantization() {
         }
       },
       "error": {
-        "message": "`.embedders.manual.binaryQuantized`: Cannot disable the binary quantization.\n - Note: Binary quantization is a lossy operation that cannot be reverted.\n - Hint: Add a new embedder that is non-quantized and regenerate the vectors.",
+        "message": "Index `doggo`: `.embedders.manual.binaryQuantized`: Cannot disable the binary quantization.\n - Note: Binary quantization is a lossy operation that cannot be reverted.\n - Hint: Add a new embedder that is non-quantized and regenerate the vectors.",
         "code": "invalid_settings_embedders",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#invalid_settings_embedders"

crates/meilisearch/tests/vector/mod.rs

@@ -250,7 +250,7 @@ async fn user_provided_embeddings_error() {
         "indexedDocuments": 0
       },
       "error": {
-        "message": "Bad embedder configuration in the document with id: `0`. Missing field `._vectors.manual.regenerate`\n  - note: `._vectors.manual` must be an array of floats, an array of arrays of floats, or an object with field `regenerate`",
+        "message": "Index `doggo`: Bad embedder configuration in the document with id: `0`. Missing field `._vectors.manual.regenerate`\n  - note: `._vectors.manual` must be an array of floats, an array of arrays of floats, or an object with field `regenerate`",
         "code": "invalid_vectors_type",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#invalid_vectors_type"
@@ -280,7 +280,7 @@ async fn user_provided_embeddings_error() {
         "indexedDocuments": 0
       },
       "error": {
-        "message": "Bad embedder configuration in the document with id: `0`. Missing field `._vectors.manual.regenerate`\n  - note: `._vectors.manual` must be an array of floats, an array of arrays of floats, or an object with field `regenerate`",
+        "message": "Index `doggo`: Bad embedder configuration in the document with id: `0`. Missing field `._vectors.manual.regenerate`\n  - note: `._vectors.manual` must be an array of floats, an array of arrays of floats, or an object with field `regenerate`",
         "code": "invalid_vectors_type",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#invalid_vectors_type"
@@ -311,7 +311,7 @@ async fn user_provided_embeddings_error() {
         "indexedDocuments": 0
       },
       "error": {
-        "message": "Bad embedder configuration in the document with id: `0`. Could not parse `._vectors.manual.regenerate`: invalid type: string \"yes please\", expected a boolean at line 1 column 26",
+        "message": "Index `doggo`: Bad embedder configuration in the document with id: `0`. Could not parse `._vectors.manual.regenerate`: invalid type: string \"yes please\", expected a boolean at line 1 column 26",
         "code": "invalid_vectors_type",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#invalid_vectors_type"
@@ -340,7 +340,7 @@ async fn user_provided_embeddings_error() {
         "indexedDocuments": 0
       },
       "error": {
-        "message": "Bad embedder configuration in the document with id: `0`. Invalid value type at `._vectors.manual.embeddings`: expected null or an array, but found a boolean: `true`",
+        "message": "Index `doggo`: Bad embedder configuration in the document with id: `0`. Invalid value type at `._vectors.manual.embeddings`: expected null or an array, but found a boolean: `true`",
         "code": "invalid_vectors_type",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#invalid_vectors_type"
@@ -369,7 +369,7 @@ async fn user_provided_embeddings_error() {
         "indexedDocuments": 0
       },
       "error": {
-        "message": "Bad embedder configuration in the document with id: `0`. Invalid value type at `._vectors.manual.embeddings[0]`: expected a number or an array, but found a boolean: `true`",
+        "message": "Index `doggo`: Bad embedder configuration in the document with id: `0`. Invalid value type at `._vectors.manual.embeddings[0]`: expected a number or an array, but found a boolean: `true`",
         "code": "invalid_vectors_type",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#invalid_vectors_type"
@@ -398,7 +398,7 @@ async fn user_provided_embeddings_error() {
         "indexedDocuments": 0
       },
       "error": {
-        "message": "Bad embedder configuration in the document with id: `0`. Invalid value type at `._vectors.manual.embeddings[0][0]`: expected a number, but found a boolean: `true`",
+        "message": "Index `doggo`: Bad embedder configuration in the document with id: `0`. Invalid value type at `._vectors.manual.embeddings[0][0]`: expected a number, but found a boolean: `true`",
         "code": "invalid_vectors_type",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#invalid_vectors_type"
@@ -440,7 +440,7 @@ async fn user_provided_embeddings_error() {
         "indexedDocuments": 0
       },
       "error": {
-        "message": "Bad embedder configuration in the document with id: `0`. Invalid value type at `._vectors.manual.embeddings[1]`: expected a number, but found an array: `[0.2,0.3]`",
+        "message": "Index `doggo`: Bad embedder configuration in the document with id: `0`. Invalid value type at `._vectors.manual.embeddings[1]`: expected a number, but found an array: `[0.2,0.3]`",
         "code": "invalid_vectors_type",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#invalid_vectors_type"
@@ -469,7 +469,7 @@ async fn user_provided_embeddings_error() {
         "indexedDocuments": 0
       },
       "error": {
-        "message": "Bad embedder configuration in the document with id: `0`. Invalid value type at `._vectors.manual.embeddings[1]`: expected an array, but found a number: `0.3`",
+        "message": "Index `doggo`: Bad embedder configuration in the document with id: `0`. Invalid value type at `._vectors.manual.embeddings[1]`: expected an array, but found a number: `0.3`",
         "code": "invalid_vectors_type",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#invalid_vectors_type"
@@ -498,7 +498,7 @@ async fn user_provided_embeddings_error() {
         "indexedDocuments": 0
       },
       "error": {
-        "message": "Bad embedder configuration in the document with id: `0`. Invalid value type at `._vectors.manual.embeddings[0][1]`: expected a number, but found a boolean: `true`",
+        "message": "Index `doggo`: Bad embedder configuration in the document with id: `0`. Invalid value type at `._vectors.manual.embeddings[0][1]`: expected a number, but found a boolean: `true`",
         "code": "invalid_vectors_type",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#invalid_vectors_type"
@@ -539,7 +539,7 @@ async fn user_provided_vectors_error() {
         "indexedDocuments": 0
       },
       "error": {
-        "message": "While embedding documents for embedder `manual`: no vectors provided for document `40` and at least 4 other document(s)\n- Note: `manual` has `source: userProvided`, so documents must provide embeddings as an array in `_vectors.manual`.\n- Hint: opt-out for a document with `_vectors.manual: null`",
+        "message": "Index `doggo`: While embedding documents for embedder `manual`: no vectors provided for document `40` and at least 4 other document(s)\n- Note: `manual` has `source: userProvided`, so documents must provide embeddings as an array in `_vectors.manual`.\n- Hint: opt-out for a document with `_vectors.manual: null`",
         "code": "vector_embedding_error",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#vector_embedding_error"
@@ -569,7 +569,7 @@ async fn user_provided_vectors_error() {
         "indexedDocuments": 0
       },
       "error": {
-        "message": "While embedding documents for embedder `manual`: no vectors provided for document `42`\n- Note: `manual` has `source: userProvided`, so documents must provide embeddings as an array in `_vectors.manual`.\n- Hint: try replacing `_vector` by `_vectors` in 1 document(s).",
+        "message": "Index `doggo`: While embedding documents for embedder `manual`: no vectors provided for document `42`\n- Note: `manual` has `source: userProvided`, so documents must provide embeddings as an array in `_vectors.manual`.\n- Hint: try replacing `_vector` by `_vectors` in 1 document(s).",
         "code": "vector_embedding_error",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#vector_embedding_error"
@@ -599,7 +599,7 @@ async fn user_provided_vectors_error() {
         "indexedDocuments": 0
       },
       "error": {
-        "message": "While embedding documents for embedder `manual`: no vectors provided for document `42`\n- Note: `manual` has `source: userProvided`, so documents must provide embeddings as an array in `_vectors.manual`.\n- Hint: try replacing `_vectors.manaul` by `_vectors.manual` in 1 document(s).",
+        "message": "Index `doggo`: While embedding documents for embedder `manual`: no vectors provided for document `42`\n- Note: `manual` has `source: userProvided`, so documents must provide embeddings as an array in `_vectors.manual`.\n- Hint: try replacing `_vectors.manaul` by `_vectors.manual` in 1 document(s).",
         "code": "vector_embedding_error",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#vector_embedding_error"

crates/meilisearch/tests/vector/openai.rs

@@ -713,7 +713,7 @@ async fn bad_api_key() {
         }
       },
       "error": {
-        "message": "While embedding documents for embedder `default`: user error: could not authenticate against OpenAI server\n  - server replied with `{\"error\":{\"message\":\"Incorrect API key provided: Bearer doggo. You can find your API key at https://platform.openai.com/account/api-keys.\",\"type\":\"invalid_request_error\",\"param\":null,\"code\":\"invalid_api_key\"}}`\n  - Hint: Check the `apiKey` parameter in the embedder configuration, and the `MEILI_OPENAI_API_KEY` and `OPENAI_API_KEY` environment variables",
+        "message": "Index `doggo`: While embedding documents for embedder `default`: user error: could not authenticate against OpenAI server\n  - server replied with `{\"error\":{\"message\":\"Incorrect API key provided: Bearer doggo. You can find your API key at https://platform.openai.com/account/api-keys.\",\"type\":\"invalid_request_error\",\"param\":null,\"code\":\"invalid_api_key\"}}`\n  - Hint: Check the `apiKey` parameter in the embedder configuration, and the `MEILI_OPENAI_API_KEY` and `OPENAI_API_KEY` environment variables",
         "code": "vector_embedding_error",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#vector_embedding_error"
@@ -757,7 +757,7 @@ async fn bad_api_key() {
         }
       },
       "error": {
-        "message": "While embedding documents for embedder `default`: user error: could not authenticate against OpenAI server\n  - server replied with `{\"error\":{\"message\":\"You didn't provide an API key. You need to provide your API key in an Authorization header using Bearer auth (i.e. Authorization: Bearer YOUR_KEY), or as the password field (with blank username) if you're accessing the API from your browser and are prompted for a username and password. You can obtain an API key from https://platform.openai.com/account/api-keys.\",\"type\":\"invalid_request_error\",\"param\":null,\"code\":null}}`\n  - Hint: Check the `apiKey` parameter in the embedder configuration, and the `MEILI_OPENAI_API_KEY` and `OPENAI_API_KEY` environment variables",
+        "message": "Index `doggo`: While embedding documents for embedder `default`: user error: could not authenticate against OpenAI server\n  - server replied with `{\"error\":{\"message\":\"You didn't provide an API key. You need to provide your API key in an Authorization header using Bearer auth (i.e. Authorization: Bearer YOUR_KEY), or as the password field (with blank username) if you're accessing the API from your browser and are prompted for a username and password. You can obtain an API key from https://platform.openai.com/account/api-keys.\",\"type\":\"invalid_request_error\",\"param\":null,\"code\":null}}`\n  - Hint: Check the `apiKey` parameter in the embedder configuration, and the `MEILI_OPENAI_API_KEY` and `OPENAI_API_KEY` environment variables",
         "code": "vector_embedding_error",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#vector_embedding_error"

crates/meilisearch/tests/vector/rest.rs

@@ -985,7 +985,7 @@ async fn bad_settings() {
         }
       },
       "error": {
-        "message": "Error while generating embeddings: runtime error: could not determine model dimensions:\n  - test embedding failed with runtime error: error extracting embeddings from the response:\n  - in `response`, while extracting a single \"{{embedding}}\", expected `response` to be an array of numbers, but failed to parse server response:\n  - invalid type: map, expected a sequence",
+        "message": "Index `doggo`: Error while generating embeddings: runtime error: could not determine model dimensions:\n  - test embedding failed with runtime error: error extracting embeddings from the response:\n  - in `response`, while extracting a single \"{{embedding}}\", expected `response` to be an array of numbers, but failed to parse server response:\n  - invalid type: map, expected a sequence",
         "code": "vector_embedding_error",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#vector_embedding_error"
@@ -1025,7 +1025,7 @@ async fn bad_settings() {
         "indexedDocuments": 0
       },
       "error": {
-        "message": "While embedding documents for embedder `rest`: runtime error: was expecting embeddings of dimension `2`, got embeddings of dimensions `3`",
+        "message": "Index `doggo`: While embedding documents for embedder `rest`: runtime error: was expecting embeddings of dimension `2`, got embeddings of dimensions `3`",
         "code": "vector_embedding_error",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#vector_embedding_error"
@@ -1178,7 +1178,7 @@ async fn server_returns_bad_request() {
         }
       },
       "error": {
-        "message": "Error while generating embeddings: runtime error: could not determine model dimensions:\n  - test embedding failed with user error: sent a bad request to embedding server\n  - Hint: check that the `request` in the embedder configuration matches the remote server's API\n  - server replied with `{\"error\":\"Invalid request: invalid type: string \\\"test\\\", expected struct MultipleRequest at line 1 column 6\"}`",
+        "message": "Index `doggo`: Error while generating embeddings: runtime error: could not determine model dimensions:\n  - test embedding failed with user error: sent a bad request to embedding server\n  - Hint: check that the `request` in the embedder configuration matches the remote server's API\n  - server replied with `{\"error\":\"Invalid request: invalid type: string \\\"test\\\", expected struct MultipleRequest at line 1 column 6\"}`",
         "code": "vector_embedding_error",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#vector_embedding_error"
@@ -1247,7 +1247,7 @@ async fn server_returns_bad_request() {
         "indexedDocuments": 0
       },
       "error": {
-        "message": "While embedding documents for embedder `rest`: user error: sent a bad request to embedding server\n  - Hint: check that the `request` in the embedder configuration matches the remote server's API\n  - server replied with `{\"error\":\"Invalid request: invalid type: string \\\"name: kefir\\\\n\\\", expected struct MultipleRequest at line 1 column 15\"}`",
+        "message": "Index `doggo`: While embedding documents for embedder `rest`: user error: sent a bad request to embedding server\n  - Hint: check that the `request` in the embedder configuration matches the remote server's API\n  - server replied with `{\"error\":\"Invalid request: invalid type: string \\\"name: kefir\\\\n\\\", expected struct MultipleRequest at line 1 column 15\"}`",
         "code": "vector_embedding_error",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#vector_embedding_error"
@@ -1306,7 +1306,7 @@ async fn server_returns_bad_response() {
         }
       },
       "error": {
-        "message": "Error while generating embeddings: runtime error: could not determine model dimensions:\n  - test embedding failed with runtime error: error extracting embeddings from the response:\n  - in `response`, while extracting the array of \"{{embedding}}\"s, configuration expects `response` to be an array with at least 1 item(s) but server sent an object with 1 field(s)",
+        "message": "Index `doggo`: Error while generating embeddings: runtime error: could not determine model dimensions:\n  - test embedding failed with runtime error: error extracting embeddings from the response:\n  - in `response`, while extracting the array of \"{{embedding}}\"s, configuration expects `response` to be an array with at least 1 item(s) but server sent an object with 1 field(s)",
         "code": "vector_embedding_error",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#vector_embedding_error"
@@ -1362,7 +1362,7 @@ async fn server_returns_bad_response() {
         }
       },
       "error": {
-        "message": "Error while generating embeddings: runtime error: could not determine model dimensions:\n  - test embedding failed with runtime error: error extracting embeddings from the response:\n  - in `response`, while extracting item #0 from the array of \"{{embedding}}\"s, expected `response` to be an array of numbers, but failed to parse server response:\n  - invalid type: map, expected a sequence",
+        "message": "Index `doggo`: Error while generating embeddings: runtime error: could not determine model dimensions:\n  - test embedding failed with runtime error: error extracting embeddings from the response:\n  - in `response`, while extracting item #0 from the array of \"{{embedding}}\"s, expected `response` to be an array of numbers, but failed to parse server response:\n  - invalid type: map, expected a sequence",
         "code": "vector_embedding_error",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#vector_embedding_error"
@@ -1414,7 +1414,7 @@ async fn server_returns_bad_response() {
         }
       },
       "error": {
-        "message": "Error while generating embeddings: runtime error: could not determine model dimensions:\n  - test embedding failed with runtime error: error extracting embeddings from the response:\n  - in `response.output`, while extracting a single \"{{embedding}}\", expected `output` to be an array of numbers, but failed to parse server response:\n  - invalid type: map, expected f32",
+        "message": "Index `doggo`: Error while generating embeddings: runtime error: could not determine model dimensions:\n  - test embedding failed with runtime error: error extracting embeddings from the response:\n  - in `response.output`, while extracting a single \"{{embedding}}\", expected `output` to be an array of numbers, but failed to parse server response:\n  - invalid type: map, expected f32",
         "code": "vector_embedding_error",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#vector_embedding_error"
@@ -1478,7 +1478,7 @@ async fn server_returns_bad_response() {
         }
       },
       "error": {
-        "message": "Error while generating embeddings: runtime error: could not determine model dimensions:\n  - test embedding failed with runtime error: error extracting embeddings from the response:\n  - in `response.embedding`, while extracting item #0 from the array of \"{{embedding}}\"s, configuration expects `embedding` to be an object with key `data` but server sent an array of size 3",
+        "message": "Index `doggo`: Error while generating embeddings: runtime error: could not determine model dimensions:\n  - test embedding failed with runtime error: error extracting embeddings from the response:\n  - in `response.embedding`, while extracting item #0 from the array of \"{{embedding}}\"s, configuration expects `embedding` to be an object with key `data` but server sent an array of size 3",
         "code": "vector_embedding_error",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#vector_embedding_error"
@@ -1542,7 +1542,7 @@ async fn server_returns_bad_response() {
         }
       },
       "error": {
-        "message": "Error while generating embeddings: runtime error: could not determine model dimensions:\n  - test embedding failed with runtime error: error extracting embeddings from the response:\n  - in `response.output[0]`, while extracting a single \"{{embedding}}\", configuration expects key \"embeddings\", which is missing in response\n  - Hint: item #0 inside `output` has key `embedding`, did you mean `response.output[0].embedding` in embedder configuration?",
+        "message": "Index `doggo`: Error while generating embeddings: runtime error: could not determine model dimensions:\n  - test embedding failed with runtime error: error extracting embeddings from the response:\n  - in `response.output[0]`, while extracting a single \"{{embedding}}\", configuration expects key \"embeddings\", which is missing in response\n  - Hint: item #0 inside `output` has key `embedding`, did you mean `response.output[0].embedding` in embedder configuration?",
         "code": "vector_embedding_error",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#vector_embedding_error"
@@ -1908,7 +1908,7 @@ async fn server_custom_header() {
         }
       },
       "error": {
-        "message": "Error while generating embeddings: runtime error: could not determine model dimensions:\n  - test embedding failed with user error: could not authenticate against embedding server\n  - server replied with `{\"error\":\"missing header 'my-nonstandard-auth'\"}`\n  - Hint: Check the `apiKey` parameter in the embedder configuration",
+        "message": "Index `doggo`: Error while generating embeddings: runtime error: could not determine model dimensions:\n  - test embedding failed with user error: could not authenticate against embedding server\n  - server replied with `{\"error\":\"missing header 'my-nonstandard-auth'\"}`\n  - Hint: Check the `apiKey` parameter in the embedder configuration",
         "code": "vector_embedding_error",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#vector_embedding_error"
@@ -1951,7 +1951,7 @@ async fn server_custom_header() {
         }
       },
       "error": {
-        "message": "Error while generating embeddings: runtime error: could not determine model dimensions:\n  - test embedding failed with user error: could not authenticate against embedding server\n  - server replied with `{\"error\":\"thou shall not pass, Balrog\"}`\n  - Hint: Check the `apiKey` parameter in the embedder configuration",
+        "message": "Index `doggo`: Error while generating embeddings: runtime error: could not determine model dimensions:\n  - test embedding failed with user error: could not authenticate against embedding server\n  - server replied with `{\"error\":\"thou shall not pass, Balrog\"}`\n  - Hint: Check the `apiKey` parameter in the embedder configuration",
         "code": "vector_embedding_error",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#vector_embedding_error"
@@ -2099,7 +2099,7 @@ async fn searchable_reindex() {
         ]
       },
       "error": {
-        "message": "While embedding documents for embedder `rest`: error: received unexpected HTTP 404 from embedding server\n  - server replied with `{\"error\":\"text not found\",\"text\":\"breed: patou\\n\"}`",
+        "message": "Index `doggo`: While embedding documents for embedder `rest`: error: received unexpected HTTP 404 from embedding server\n  - server replied with `{\"error\":\"text not found\",\"text\":\"breed: patou\\n\"}`",
         "code": "vector_embedding_error",
         "type": "invalid_request",
         "link": "https://docs.meilisearch.com/errors#vector_embedding_error"

crates/milli/Cargo.toml

@@ -98,7 +98,7 @@ allocator-api2 = "0.2.18"
 rustc-hash = "2.0.0"
 uell = "0.1.0"
 enum-iterator = "2.1.0"
-bbqueue = { git = "https://github.com/kerollmops/bbqueue" }
+bbqueue = { git = "https://github.com/meilisearch/bbqueue" }
 flume = { version = "0.11.1", default-features = false }
 
 [dev-dependencies]

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

@@ -7,7 +7,7 @@ use std::num::NonZeroU16;
 use std::ops::Range;
 use std::time::Duration;
 
-use bbqueue::framed::{FrameGrantR, FrameGrantW, FrameProducer};
+use bbqueue::framed::{FrameGrantR, FrameProducer};
 use bbqueue::BBBuffer;
 use bytemuck::{checked, CheckedBitPattern, NoUninit};
 use flume::{RecvTimeoutError, SendError};
@@ -21,11 +21,15 @@ use super::ref_cell_ext::RefCellExt;
 use super::thread_local::{FullySend, ThreadLocal};
 use super::StdResult;
 use crate::heed_codec::facet::{FieldDocIdFacetF64Codec, FieldDocIdFacetStringCodec};
+use crate::heed_codec::StrBEU16Codec;
 use crate::index::db_name;
 use crate::index::main_key::{GEO_FACETED_DOCUMENTS_IDS_KEY, GEO_RTREE_KEY};
 use crate::update::new::KvReaderFieldId;
 use crate::vector::Embedding;
-use crate::{CboRoaringBitmapCodec, DocumentId, Error, Index, InternalError};
+use crate::{
+    CboRoaringBitmapCodec, DocumentId, Error, FieldIdWordCountCodec, Index, InternalError,
+    U8StrStrCodec,
+};
 
 /// Creates a tuple of senders/receiver to be used by
 /// the extractors and the writer loop.
@@ -403,6 +407,32 @@ impl Database {
             Database::FieldIdDocidFacetF64s => db_name::FIELD_ID_DOCID_FACET_F64S,
         }
     }
+
+    pub fn stringify_key(&self, key: &[u8]) -> String {
+        use heed::types::*;
+
+        match self {
+            Database::WordDocids => format!("{:?}", Str::bytes_decode(key).unwrap()),
+            Database::WordFidDocids => format!("{:?}", StrBEU16Codec::bytes_decode(key).unwrap()),
+            Database::WordPositionDocids => {
+                format!("{:?}", StrBEU16Codec::bytes_decode(key).unwrap())
+            }
+            Database::WordPairProximityDocids => {
+                format!("{:?}", U8StrStrCodec::bytes_decode(key).unwrap())
+            }
+            Database::ExactWordDocids => format!("{:?}", Str::bytes_decode(key).unwrap()),
+            Database::FidWordCountDocids => {
+                format!("{:?}", FieldIdWordCountCodec::bytes_decode(key).unwrap())
+            }
+            Database::FieldIdDocidFacetStrings => {
+                format!("{:?}", FieldDocIdFacetStringCodec::bytes_decode(key).unwrap())
+            }
+            Database::FieldIdDocidFacetF64s => {
+                format!("{:?}", FieldDocIdFacetF64Codec::bytes_decode(key).unwrap())
+            }
+            d => unimplemented!("stringify_key for {:?}", d),
+        }
+    }
 }
 
 impl From<FacetKind> for Database {
@@ -454,14 +484,10 @@ impl<'b> ExtractorBbqueueSender<'b> {
         }
 
         // Spin loop to have a frame the size we requested.
-        let mut grant = reserve_grant(&mut producer, total_length, &self.sender)?;
-        payload_header.serialize_into(&mut grant);
-
-        // We only send a wake up message when the channel is empty
-        // so that we don't fill the channel with too many WakeUps.
-        if self.sender.is_empty() {
-            self.sender.send(ReceiverAction::WakeUp).unwrap();
-        }
+        reserve_and_write_grant(&mut producer, total_length, &self.sender, |grant| {
+            payload_header.serialize_into(grant);
+            Ok(())
+        })?;
 
         Ok(())
     }
@@ -500,24 +526,20 @@ impl<'b> ExtractorBbqueueSender<'b> {
         }
 
         // Spin loop to have a frame the size we requested.
-        let mut grant = reserve_grant(&mut producer, total_length, &self.sender)?;
+        reserve_and_write_grant(&mut producer, total_length, &self.sender, |grant| {
+            let header_size = payload_header.header_size();
+            let (header_bytes, remaining) = grant.split_at_mut(header_size);
+            payload_header.serialize_into(header_bytes);
 
-        let header_size = payload_header.header_size();
-        let (header_bytes, remaining) = grant.split_at_mut(header_size);
-        payload_header.serialize_into(header_bytes);
-
-        if dimensions != 0 {
-            let output_iter = remaining.chunks_exact_mut(dimensions * mem::size_of::<f32>());
-            for (embedding, output) in embeddings.iter().zip(output_iter) {
-                output.copy_from_slice(bytemuck::cast_slice(embedding));
+            if dimensions != 0 {
+                let output_iter = remaining.chunks_exact_mut(dimensions * mem::size_of::<f32>());
+                for (embedding, output) in embeddings.iter().zip(output_iter) {
+                    output.copy_from_slice(bytemuck::cast_slice(embedding));
+                }
             }
-        }
 
-        // We only send a wake up message when the channel is empty
-        // so that we don't fill the channel with too many WakeUps.
-        if self.sender.is_empty() {
-            self.sender.send(ReceiverAction::WakeUp).unwrap();
-        }
+            Ok(())
+        })?;
 
         Ok(())
     }
@@ -575,19 +597,13 @@ impl<'b> ExtractorBbqueueSender<'b> {
         }
 
         // Spin loop to have a frame the size we requested.
-        let mut grant = reserve_grant(&mut producer, total_length, &self.sender)?;
-
-        let header_size = payload_header.header_size();
-        let (header_bytes, remaining) = grant.split_at_mut(header_size);
-        payload_header.serialize_into(header_bytes);
-        let (key_buffer, value_buffer) = remaining.split_at_mut(key_length.get() as usize);
-        key_value_writer(key_buffer, value_buffer)?;
-
-        // We only send a wake up message when the channel is empty
-        // so that we don't fill the channel with too many WakeUps.
-        if self.sender.is_empty() {
-            self.sender.send(ReceiverAction::WakeUp).unwrap();
-        }
+        reserve_and_write_grant(&mut producer, total_length, &self.sender, |grant| {
+            let header_size = payload_header.header_size();
+            let (header_bytes, remaining) = grant.split_at_mut(header_size);
+            payload_header.serialize_into(header_bytes);
+            let (key_buffer, value_buffer) = remaining.split_at_mut(key_length.get() as usize);
+            key_value_writer(key_buffer, value_buffer)
+        })?;
 
         Ok(())
     }
@@ -629,37 +645,44 @@ impl<'b> ExtractorBbqueueSender<'b> {
         }
 
         // Spin loop to have a frame the size we requested.
-        let mut grant = reserve_grant(&mut producer, total_length, &self.sender)?;
-
-        let header_size = payload_header.header_size();
-        let (header_bytes, remaining) = grant.split_at_mut(header_size);
-        payload_header.serialize_into(header_bytes);
-        key_writer(remaining)?;
-
-        // We only send a wake up message when the channel is empty
-        // so that we don't fill the channel with too many WakeUps.
-        if self.sender.is_empty() {
-            self.sender.send(ReceiverAction::WakeUp).unwrap();
-        }
+        reserve_and_write_grant(&mut producer, total_length, &self.sender, |grant| {
+            let header_size = payload_header.header_size();
+            let (header_bytes, remaining) = grant.split_at_mut(header_size);
+            payload_header.serialize_into(header_bytes);
+            key_writer(remaining)
+        })?;
 
         Ok(())
     }
 }
 
-/// Try to reserve a frame grant of `total_length` by spin looping
-/// on the BBQueue buffer and panics if the receiver has been disconnected.
-fn reserve_grant<'b>(
-    producer: &mut FrameProducer<'b>,
+/// Try to reserve a frame grant of `total_length` by spin
+/// looping on the BBQueue buffer, panics if the receiver
+/// has been disconnected or send a WakeUp message if necessary.
+fn reserve_and_write_grant<F>(
+    producer: &mut FrameProducer,
     total_length: usize,
     sender: &flume::Sender<ReceiverAction>,
-) -> crate::Result<FrameGrantW<'b>> {
+    f: F,
+) -> crate::Result<()>
+where
+    F: FnOnce(&mut [u8]) -> crate::Result<()>,
+{
     loop {
         for _ in 0..10_000 {
             match producer.grant(total_length) {
                 Ok(mut grant) => {
                     // We could commit only the used memory.
-                    grant.to_commit(total_length);
-                    return Ok(grant);
+                    f(&mut grant)?;
+                    grant.commit(total_length);
+
+                    // We only send a wake up message when the channel is empty
+                    // so that we don't fill the channel with too many WakeUps.
+                    if sender.is_empty() {
+                        sender.send(ReceiverAction::WakeUp).unwrap();
+                    }
+
+                    return Ok(());
                 }
                 Err(bbqueue::Error::InsufficientSize) => continue,
                 Err(e) => unreachable!("{e:?}"),
@@ -668,6 +691,11 @@ fn reserve_grant<'b>(
         if sender.is_disconnected() {
             return Err(Error::InternalError(InternalError::AbortedIndexation));
         }
+
+        // We prefer to yield and allow the writing thread
+        // to do its job, especially beneficial when there
+        // is only one CPU core available.
+        std::thread::yield_now();
     }
 }
 

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

@@ -1,7 +1,10 @@
 use bumpalo::Bump;
 use heed::RoTxn;
 
-use super::document::{DocumentFromDb, DocumentFromVersions, MergedDocument, Versions};
+use super::document::{
+    Document as _, DocumentFromDb, DocumentFromVersions, MergedDocument, Versions,
+};
+use super::extract::perm_json_p;
 use super::vector_document::{
     MergedVectorDocument, VectorDocumentFromDb, VectorDocumentFromVersions,
 };
@@ -164,6 +167,80 @@ impl<'doc> Update<'doc> {
         }
     }
 
+    /// Returns whether the updated version of the document is different from the current version for the passed subset of fields.
+    ///
+    /// `true` if at least one top-level-field that is a exactly a member of field or a parent of a member of field changed.
+    /// Otherwise `false`.
+    pub fn has_changed_for_fields<'t, Mapper: FieldIdMapper>(
+        &self,
+        fields: Option<&[&str]>,
+        rtxn: &'t RoTxn,
+        index: &'t Index,
+        mapper: &'t Mapper,
+    ) -> Result<bool> {
+        let mut changed = false;
+        let mut cached_current = None;
+        let mut updated_selected_field_count = 0;
+
+        for entry in self.updated().iter_top_level_fields() {
+            let (key, updated_value) = entry?;
+
+            if perm_json_p::select_field(key, fields, &[]) == perm_json_p::Selection::Skip {
+                continue;
+            }
+
+            updated_selected_field_count += 1;
+            let current = match cached_current {
+                Some(current) => current,
+                None => self.current(rtxn, index, mapper)?,
+            };
+            let current_value = current.top_level_field(key)?;
+            let Some(current_value) = current_value else {
+                changed = true;
+                break;
+            };
+
+            if current_value.get() != updated_value.get() {
+                changed = true;
+                break;
+            }
+            cached_current = Some(current);
+        }
+
+        if !self.has_deletion {
+            // no field deletion, so fields that don't appear in `updated` cannot have changed
+            return Ok(changed);
+        }
+
+        if changed {
+            return Ok(true);
+        }
+
+        // we saw all updated fields, and set `changed` if any field wasn't in `current`.
+        // so if there are as many fields in `current` as in `updated`, then nothing changed.
+        // If there is any more fields in `current`, then they are missing in `updated`.
+        let has_deleted_fields = {
+            let current = match cached_current {
+                Some(current) => current,
+                None => self.current(rtxn, index, mapper)?,
+            };
+
+            let mut current_selected_field_count = 0;
+            for entry in current.iter_top_level_fields() {
+                let (key, _) = entry?;
+
+                if perm_json_p::select_field(key, fields, &[]) == perm_json_p::Selection::Skip {
+                    continue;
+                }
+                current_selected_field_count += 1;
+            }
+
+            current_selected_field_count != updated_selected_field_count
+        };
+
+        Ok(has_deleted_fields)
+    }
+
     pub fn updated_vectors(
         &self,
         doc_alloc: &'doc Bump,

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

@@ -466,12 +466,13 @@ pub fn transpose_and_freeze_caches<'a, 'extractor>(
     Ok(bucket_caches)
 }
 
-/// Merges the caches that must be all associated to the same bucket.
+/// Merges the caches that must be all associated to the same bucket
+/// but make sure to sort the different buckets before performing the merges.
 ///
 /// # Panics
 ///
 /// - If the bucket IDs in these frozen caches are not exactly the same.
-pub fn merge_caches<F>(frozen: Vec<FrozenCache>, mut f: F) -> Result<()>
+pub fn merge_caches_sorted<F>(frozen: Vec<FrozenCache>, mut f: F) -> Result<()>
 where
     F: for<'a> FnMut(&'a [u8], DelAddRoaringBitmap) -> Result<()>,
 {
@@ -543,12 +544,12 @@ where
 
     // Then manage the content on the HashMap entries that weren't taken (mem::take).
     while let Some(mut map) = maps.pop() {
-        for (key, bbbul) in map.iter_mut() {
-            // Make sure we don't try to work with entries already managed by the spilled
-            if bbbul.is_empty() {
-                continue;
-            }
+        // Make sure we don't try to work with entries already managed by the spilled
+        let mut ordered_entries: Vec<_> =
+            map.iter_mut().filter(|(_, bbbul)| !bbbul.is_empty()).collect();
+        ordered_entries.sort_unstable_by_key(|(key, _)| *key);
 
+        for (key, bbbul) in ordered_entries {
             let mut output = DelAddRoaringBitmap::empty();
             output.union_and_clear_bbbul(bbbul);
 

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

@@ -97,6 +97,15 @@ impl FacetedDocidsExtractor {
                 },
             ),
             DocumentChange::Update(inner) => {
+                if !inner.has_changed_for_fields(
+                    Some(attributes_to_extract),
+                    rtxn,
+                    index,
+                    context.db_fields_ids_map,
+                )? {
+                    return Ok(());
+                }
+
                 extract_document_facets(
                     attributes_to_extract,
                     inner.current(rtxn, index, context.db_fields_ids_map)?,

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

@@ -1,6 +1,6 @@
 use std::cell::RefCell;
 use std::fs::File;
-use std::io::{self, BufReader, BufWriter, ErrorKind, Read, Write as _};
+use std::io::{self, BufReader, BufWriter, ErrorKind, Read, Seek as _, Write as _};
 use std::{iter, mem, result};
 
 use bumpalo::Bump;
@@ -97,30 +97,34 @@ pub struct FrozenGeoExtractorData<'extractor> {
 impl<'extractor> FrozenGeoExtractorData<'extractor> {
     pub fn iter_and_clear_removed(
         &mut self,
-    ) -> impl IntoIterator<Item = io::Result<ExtractedGeoPoint>> + '_ {
-        mem::take(&mut self.removed)
+    ) -> io::Result<impl IntoIterator<Item = io::Result<ExtractedGeoPoint>> + '_> {
+        Ok(mem::take(&mut self.removed)
             .iter()
             .copied()
             .map(Ok)
-            .chain(iterator_over_spilled_geopoints(&mut self.spilled_removed))
+            .chain(iterator_over_spilled_geopoints(&mut self.spilled_removed)?))
     }
 
     pub fn iter_and_clear_inserted(
         &mut self,
-    ) -> impl IntoIterator<Item = io::Result<ExtractedGeoPoint>> + '_ {
-        mem::take(&mut self.inserted)
+    ) -> io::Result<impl IntoIterator<Item = io::Result<ExtractedGeoPoint>> + '_> {
+        Ok(mem::take(&mut self.inserted)
             .iter()
             .copied()
             .map(Ok)
-            .chain(iterator_over_spilled_geopoints(&mut self.spilled_inserted))
+            .chain(iterator_over_spilled_geopoints(&mut self.spilled_inserted)?))
     }
 }
 
 fn iterator_over_spilled_geopoints(
     spilled: &mut Option<BufReader<File>>,
-) -> impl IntoIterator<Item = io::Result<ExtractedGeoPoint>> + '_ {
+) -> io::Result<impl IntoIterator<Item = io::Result<ExtractedGeoPoint>> + '_> {
     let mut spilled = spilled.take();
-    iter::from_fn(move || match &mut spilled {
+    if let Some(spilled) = &mut spilled {
+        spilled.rewind()?;
+    }
+
+    Ok(iter::from_fn(move || match &mut spilled {
         Some(file) => {
             let geopoint_bytes = &mut [0u8; mem::size_of::<ExtractedGeoPoint>()];
             match file.read_exact(geopoint_bytes) {
@@ -130,7 +134,7 @@ fn iterator_over_spilled_geopoints(
             }
         }
         None => None,
-    })
+    }))
 }
 
 impl<'extractor> Extractor<'extractor> for GeoExtractor {
@@ -157,7 +161,9 @@ impl<'extractor> Extractor<'extractor> for GeoExtractor {
         let mut data_ref = context.data.borrow_mut_or_yield();
 
         for change in changes {
-            if max_memory.map_or(false, |mm| context.extractor_alloc.allocated_bytes() >= mm) {
+            if data_ref.spilled_removed.is_none()
+                && max_memory.map_or(false, |mm| context.extractor_alloc.allocated_bytes() >= mm)
+            {
                 // We must spill as we allocated too much memory
                 data_ref.spilled_removed = tempfile::tempfile().map(BufWriter::new).map(Some)?;
                 data_ref.spilled_inserted = tempfile::tempfile().map(BufWriter::new).map(Some)?;

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

@@ -6,7 +6,9 @@ mod searchable;
 mod vectors;
 
 use bumpalo::Bump;
-pub use cache::{merge_caches, transpose_and_freeze_caches, BalancedCaches, DelAddRoaringBitmap};
+pub use cache::{
+    merge_caches_sorted, transpose_and_freeze_caches, BalancedCaches, DelAddRoaringBitmap,
+};
 pub use documents::*;
 pub use faceted::*;
 pub use geo::*;

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

@@ -28,7 +28,7 @@ pub struct WordDocidsBalancedCaches<'extractor> {
     exact_word_docids: BalancedCaches<'extractor>,
     word_position_docids: BalancedCaches<'extractor>,
     fid_word_count_docids: BalancedCaches<'extractor>,
-    fid_word_count: HashMap<FieldId, (usize, usize)>,
+    fid_word_count: HashMap<FieldId, (Option<usize>, Option<usize>)>,
     current_docid: Option<DocumentId>,
 }
 
@@ -85,8 +85,8 @@ impl<'extractor> WordDocidsBalancedCaches<'extractor> {
 
         self.fid_word_count
             .entry(field_id)
-            .and_modify(|(_current_count, new_count)| *new_count += 1)
-            .or_insert((0, 1));
+            .and_modify(|(_current_count, new_count)| *new_count.get_or_insert(0) += 1)
+            .or_insert((None, Some(1)));
         self.current_docid = Some(docid);
 
         Ok(())
@@ -130,8 +130,8 @@ impl<'extractor> WordDocidsBalancedCaches<'extractor> {
 
         self.fid_word_count
             .entry(field_id)
-            .and_modify(|(current_count, _new_count)| *current_count += 1)
-            .or_insert((1, 0));
+            .and_modify(|(current_count, _new_count)| *current_count.get_or_insert(0) += 1)
+            .or_insert((Some(1), None));
 
         self.current_docid = Some(docid);
 
@@ -141,14 +141,18 @@ impl<'extractor> WordDocidsBalancedCaches<'extractor> {
     fn flush_fid_word_count(&mut self, buffer: &mut BumpVec<u8>) -> Result<()> {
         for (fid, (current_count, new_count)) in self.fid_word_count.drain() {
             if current_count != new_count {
-                if current_count <= MAX_COUNTED_WORDS {
+                if let Some(current_count) =
+                    current_count.filter(|current_count| *current_count <= MAX_COUNTED_WORDS)
+                {
                     buffer.clear();
                     buffer.extend_from_slice(&fid.to_be_bytes());
                     buffer.push(current_count as u8);
                     self.fid_word_count_docids
                         .insert_del_u32(buffer, self.current_docid.unwrap())?;
                 }
-                if new_count <= MAX_COUNTED_WORDS {
+                if let Some(new_count) =
+                    new_count.filter(|new_count| *new_count <= MAX_COUNTED_WORDS)
+                {
                     buffer.clear();
                     buffer.extend_from_slice(&fid.to_be_bytes());
                     buffer.push(new_count as u8);
@@ -351,6 +355,15 @@ impl WordDocidsExtractors {
                 )?;
             }
             DocumentChange::Update(inner) => {
+                if !inner.has_changed_for_fields(
+                    document_tokenizer.attribute_to_extract,
+                    &context.rtxn,
+                    context.index,
+                    context.db_fields_ids_map,
+                )? {
+                    return Ok(());
+                }
+
                 let mut token_fn = |fname: &str, fid, pos, word: &str| {
                     cached_sorter.insert_del_u32(
                         fid,

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

@@ -70,6 +70,15 @@ impl SearchableExtractor for WordPairProximityDocidsExtractor {
                 )?;
             }
             DocumentChange::Update(inner) => {
+                if !inner.has_changed_for_fields(
+                    document_tokenizer.attribute_to_extract,
+                    rtxn,
+                    index,
+                    context.db_fields_ids_map,
+                )? {
+                    return Ok(());
+                }
+
                 let document = inner.current(rtxn, index, context.db_fields_ids_map)?;
                 process_document_tokens(
                     document,

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

@@ -80,15 +80,26 @@ where
     let mut bbbuffers = Vec::new();
     let finished_extraction = AtomicBool::new(false);
 
+    // We reduce the actual memory used to 5%. The reason we do this here and not in Meilisearch
+    // is because we still use the old indexer for the settings and it is highly impacted by the
+    // max memory. So we keep the changes here and will remove these changes once we use the new
+    // indexer to also index settings. Related to #5125 and #5141.
+    let grenad_parameters = GrenadParameters {
+        max_memory: grenad_parameters.max_memory.map(|mm| mm * 5 / 100),
+        ..grenad_parameters
+    };
+
     // We compute and remove the allocated BBQueues buffers capacity from the indexing memory.
     let minimum_capacity = 50 * 1024 * 1024 * pool.current_num_threads(); // 50 MiB
     let (grenad_parameters, total_bbbuffer_capacity) = grenad_parameters.max_memory.map_or(
         (grenad_parameters, 2 * minimum_capacity), // 100 MiB by thread by default
         |max_memory| {
             // 2% of the indexing memory
-            let total_bbbuffer_capacity = (max_memory / 100 / 2).min(minimum_capacity);
+            let total_bbbuffer_capacity = (max_memory / 100 / 2).max(minimum_capacity);
             let new_grenad_parameters = GrenadParameters {
-                max_memory: Some(max_memory - total_bbbuffer_capacity),
+                max_memory: Some(
+                    max_memory.saturating_sub(total_bbbuffer_capacity).max(100 * 1024 * 1024),
+                ),
                 ..grenad_parameters
             };
             (new_grenad_parameters, total_bbbuffer_capacity)

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

@@ -9,8 +9,8 @@ use roaring::RoaringBitmap;
 
 use super::channel::*;
 use super::extract::{
-    merge_caches, transpose_and_freeze_caches, BalancedCaches, DelAddRoaringBitmap, FacetKind,
-    GeoExtractorData,
+    merge_caches_sorted, transpose_and_freeze_caches, BalancedCaches, DelAddRoaringBitmap,
+    FacetKind, GeoExtractorData,
 };
 use crate::{CboRoaringBitmapCodec, FieldId, GeoPoint, Index, InternalError, Result};
 
@@ -34,7 +34,7 @@ where
         }
 
         let mut frozen = data.into_inner().freeze()?;
-        for result in frozen.iter_and_clear_removed() {
+        for result in frozen.iter_and_clear_removed()? {
             let extracted_geo_point = result?;
             let removed = rtree.remove(&GeoPoint::from(extracted_geo_point));
             debug_assert!(removed.is_some());
@@ -42,7 +42,7 @@ where
             debug_assert!(removed);
         }
 
-        for result in frozen.iter_and_clear_inserted() {
+        for result in frozen.iter_and_clear_inserted()? {
             let extracted_geo_point = result?;
             rtree.insert(GeoPoint::from(extracted_geo_point));
             let inserted = faceted.insert(extracted_geo_point.docid);
@@ -78,18 +78,37 @@ where
         if must_stop_processing() {
             return Err(InternalError::AbortedIndexation.into());
         }
-        merge_caches(frozen, |key, DelAddRoaringBitmap { del, add }| {
+        merge_caches_sorted(frozen, |key, DelAddRoaringBitmap { del, add }| {
             let current = database.get(&rtxn, key)?;
-            match merge_cbo_bitmaps(current, del, add)? {
-                Operation::Write(bitmap) => {
+
+            if let (Some(del), Some(current)) = (&del, &current) {
+                let current = CboRoaringBitmapCodec::deserialize_from(current).unwrap();
+                let diff = del - &current;
+                let external_ids = index.external_id_of(&rtxn, &diff).unwrap().into_iter().map(|id| id.unwrap()).collect::<Vec<_>>();
+                if !del.is_subset(&current) {
+                    eprintln!(
+                        "======================== {:?}: {} ->  c: {:?} d: {:?} a: {:?} extra: {:?} extra_external_ids: {:?}",
+                        D::DATABASE,
+                        D::DATABASE.stringify_key(key),
+                        &current,
+                        del,
+                        add,
+                        diff,
+                        external_ids
+                    );
+                }
+            }
+            match merge_cbo_bitmaps(current, del, add) {
+                Ok(Operation::Write(bitmap)) => {
                     docids_sender.write(key, &bitmap)?;
                     Ok(())
                 }
-                Operation::Delete => {
+                Ok(Operation::Delete) => {
                     docids_sender.delete(key)?;
                     Ok(())
                 }
-                Operation::Ignore => Ok(()),
+                Ok(Operation::Ignore) => Ok(()),
+                Err(e) => Err(e),
             }
         })
     })
@@ -107,7 +126,7 @@ pub fn merge_and_send_facet_docids<'extractor>(
         .map(|frozen| {
             let mut facet_field_ids_delta = FacetFieldIdsDelta::default();
             let rtxn = index.read_txn()?;
-            merge_caches(frozen, |key, DelAddRoaringBitmap { del, add }| {
+            merge_caches_sorted(frozen, |key, DelAddRoaringBitmap { del, add }| {
                 let current = database.get_cbo_roaring_bytes_value(&rtxn, key)?;
                 match merge_cbo_bitmaps(current, del, add)? {
                     Operation::Write(bitmap) => {
@@ -234,17 +253,46 @@ fn merge_cbo_bitmaps(
         (None, Some(_del), Some(add)) => Ok(Operation::Write(add)),
         (Some(_current), None, None) => Ok(Operation::Ignore), // but it's strange
         (Some(current), None, Some(add)) => Ok(Operation::Write(current | add)),
-        (Some(current), Some(del), add) => {
+        (Some(current), Some(mut del), add) => {
+            debug_assert!(
+                del.is_subset(&current),
+                "del is not a subset of current, which must be impossible."
+            );
             let output = match add {
-                Some(add) => (&current - del) | add,
-                None => &current - del,
+                Some(add) => {
+                    del -= &add;
+
+                    if del.is_empty() {
+                        if add.is_subset(&current) {
+                            // no changes, no allocation
+                            None
+                        } else {
+                            // addition
+                            Some(current | add)
+                        }
+                    } else {
+                        if add.is_subset(&current) {
+                            // deletion only, no union
+                            Some(current - del)
+                        } else {
+                            // deletion and addition
+                            Some((current - del) | add)
+                        }
+                    }
+                }
+                // deletion only, no union
+                None => Some(current - del),
             };
-            if output.is_empty() {
-                Ok(Operation::Delete)
-            } else if current == output {
-                Ok(Operation::Ignore)
-            } else {
-                Ok(Operation::Write(output))
+
+            match output {
+                Some(output) => {
+                    if output.is_empty() {
+                        Ok(Operation::Delete)
+                    } else {
+                        Ok(Operation::Write(output))
+                    }
+                }
+                None => Ok(Operation::Ignore),
             }
         }
     }

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

@@ -1,4 +1,4 @@
-use std::collections::HashSet;
+use std::collections::BTreeSet;
 use std::io::BufWriter;
 
 use fst::{Set, SetBuilder, Streamer};
@@ -75,8 +75,8 @@ pub struct PrefixData {
 
 #[derive(Debug)]
 pub struct PrefixDelta {
-    pub modified: HashSet<Prefix>,
-    pub deleted: HashSet<Prefix>,
+    pub modified: BTreeSet<Prefix>,
+    pub deleted: BTreeSet<Prefix>,
 }
 
 struct PrefixFstBuilder {
@@ -86,7 +86,7 @@ struct PrefixFstBuilder {
     prefix_fst_builders: Vec<SetBuilder<Vec<u8>>>,
     current_prefix: Vec<Prefix>,
     current_prefix_count: Vec<usize>,
-    modified_prefixes: HashSet<Prefix>,
+    modified_prefixes: BTreeSet<Prefix>,
     current_prefix_is_modified: Vec<bool>,
 }
 
@@ -110,7 +110,7 @@ impl PrefixFstBuilder {
             prefix_fst_builders,
             current_prefix: vec![Prefix::new(); max_prefix_length],
             current_prefix_count: vec![0; max_prefix_length],
-            modified_prefixes: HashSet::new(),
+            modified_prefixes: BTreeSet::new(),
             current_prefix_is_modified: vec![false; max_prefix_length],
         })
     }
@@ -180,7 +180,7 @@ impl PrefixFstBuilder {
         let prefix_fst_mmap = unsafe { Mmap::map(&prefix_fst_file)? };
         let new_prefix_fst = Set::new(&prefix_fst_mmap)?;
         let old_prefix_fst = index.words_prefixes_fst(rtxn)?;
-        let mut deleted_prefixes = HashSet::new();
+        let mut deleted_prefixes = BTreeSet::new();
         {
             let mut deleted_prefixes_stream = old_prefix_fst.op().add(&new_prefix_fst).difference();
             while let Some(prefix) = deleted_prefixes_stream.next() {

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

@@ -1,5 +1,5 @@
 use std::cell::RefCell;
-use std::collections::HashSet;
+use std::collections::BTreeSet;
 use std::io::{BufReader, BufWriter, Read, Seek, Write};
 
 use hashbrown::HashMap;
@@ -37,8 +37,8 @@ impl WordPrefixDocids {
     fn execute(
         self,
         wtxn: &mut heed::RwTxn,
-        prefix_to_compute: &HashSet<Prefix>,
-        prefix_to_delete: &HashSet<Prefix>,
+        prefix_to_compute: &BTreeSet<Prefix>,
+        prefix_to_delete: &BTreeSet<Prefix>,
     ) -> Result<()> {
         delete_prefixes(wtxn, &self.prefix_database, prefix_to_delete)?;
         self.recompute_modified_prefixes(wtxn, prefix_to_compute)
@@ -48,7 +48,7 @@ impl WordPrefixDocids {
     fn recompute_modified_prefixes(
         &self,
         wtxn: &mut RwTxn,
-        prefixes: &HashSet<Prefix>,
+        prefixes: &BTreeSet<Prefix>,
     ) -> Result<()> {
         // We fetch the docids associated to the newly added word prefix fst only.
         // And collect the CboRoaringBitmaps pointers in an HashMap.
@@ -127,7 +127,7 @@ impl<'a, 'rtxn> FrozenPrefixBitmaps<'a, 'rtxn> {
     pub fn from_prefixes(
         database: Database<Bytes, CboRoaringBitmapCodec>,
         rtxn: &'rtxn RoTxn,
-        prefixes: &'a HashSet<Prefix>,
+        prefixes: &'a BTreeSet<Prefix>,
     ) -> heed::Result<Self> {
         let database = database.remap_data_type::<Bytes>();
 
@@ -173,8 +173,8 @@ impl WordPrefixIntegerDocids {
     fn execute(
         self,
         wtxn: &mut heed::RwTxn,
-        prefix_to_compute: &HashSet<Prefix>,
-        prefix_to_delete: &HashSet<Prefix>,
+        prefix_to_compute: &BTreeSet<Prefix>,
+        prefix_to_delete: &BTreeSet<Prefix>,
     ) -> Result<()> {
         delete_prefixes(wtxn, &self.prefix_database, prefix_to_delete)?;
         self.recompute_modified_prefixes(wtxn, prefix_to_compute)
@@ -184,7 +184,7 @@ impl WordPrefixIntegerDocids {
     fn recompute_modified_prefixes(
         &self,
         wtxn: &mut RwTxn,
-        prefixes: &HashSet<Prefix>,
+        prefixes: &BTreeSet<Prefix>,
     ) -> Result<()> {
         // We fetch the docids associated to the newly added word prefix fst only.
         // And collect the CboRoaringBitmaps pointers in an HashMap.
@@ -262,7 +262,7 @@ impl<'a, 'rtxn> FrozenPrefixIntegerBitmaps<'a, 'rtxn> {
     pub fn from_prefixes(
         database: Database<Bytes, CboRoaringBitmapCodec>,
         rtxn: &'rtxn RoTxn,
-        prefixes: &'a HashSet<Prefix>,
+        prefixes: &'a BTreeSet<Prefix>,
     ) -> heed::Result<Self> {
         let database = database.remap_data_type::<Bytes>();
 
@@ -291,7 +291,7 @@ unsafe impl<'a, 'rtxn> Sync for FrozenPrefixIntegerBitmaps<'a, 'rtxn> {}
 fn delete_prefixes(
     wtxn: &mut RwTxn,
     prefix_database: &Database<Bytes, CboRoaringBitmapCodec>,
-    prefixes: &HashSet<Prefix>,
+    prefixes: &BTreeSet<Prefix>,
 ) -> Result<()> {
     // We remove all the entries that are no more required in this word prefix docids database.
     for prefix in prefixes {
@@ -309,8 +309,8 @@ fn delete_prefixes(
 pub fn compute_word_prefix_docids(
     wtxn: &mut RwTxn,
     index: &Index,
-    prefix_to_compute: &HashSet<Prefix>,
-    prefix_to_delete: &HashSet<Prefix>,
+    prefix_to_compute: &BTreeSet<Prefix>,
+    prefix_to_delete: &BTreeSet<Prefix>,
     grenad_parameters: GrenadParameters,
 ) -> Result<()> {
     WordPrefixDocids::new(
@@ -325,8 +325,8 @@ pub fn compute_word_prefix_docids(
 pub fn compute_exact_word_prefix_docids(
     wtxn: &mut RwTxn,
     index: &Index,
-    prefix_to_compute: &HashSet<Prefix>,
-    prefix_to_delete: &HashSet<Prefix>,
+    prefix_to_compute: &BTreeSet<Prefix>,
+    prefix_to_delete: &BTreeSet<Prefix>,
     grenad_parameters: GrenadParameters,
 ) -> Result<()> {
     WordPrefixDocids::new(
@@ -341,8 +341,8 @@ pub fn compute_exact_word_prefix_docids(
 pub fn compute_word_prefix_fid_docids(
     wtxn: &mut RwTxn,
     index: &Index,
-    prefix_to_compute: &HashSet<Prefix>,
-    prefix_to_delete: &HashSet<Prefix>,
+    prefix_to_compute: &BTreeSet<Prefix>,
+    prefix_to_delete: &BTreeSet<Prefix>,
     grenad_parameters: GrenadParameters,
 ) -> Result<()> {
     WordPrefixIntegerDocids::new(
@@ -357,8 +357,8 @@ pub fn compute_word_prefix_fid_docids(
 pub fn compute_word_prefix_position_docids(
     wtxn: &mut RwTxn,
     index: &Index,
-    prefix_to_compute: &HashSet<Prefix>,
-    prefix_to_delete: &HashSet<Prefix>,
+    prefix_to_compute: &BTreeSet<Prefix>,
+    prefix_to_delete: &BTreeSet<Prefix>,
     grenad_parameters: GrenadParameters,
 ) -> Result<()> {
     WordPrefixIntegerDocids::new(

crates/xtask/src/bench/mod.rs

@@ -82,6 +82,10 @@ pub struct BenchDeriveArgs {
     /// Reason for the benchmark invocation
     #[arg(short, long)]
     reason: Option<String>,
+
+    /// The maximum time in seconds we allow for fetching the task queue before timing out.
+    #[arg(long, default_value_t = 60)]
+    tasks_queue_timeout_secs: u64,
 }
 
 pub fn run(args: BenchDeriveArgs) -> anyhow::Result<()> {
@@ -127,7 +131,7 @@ pub fn run(args: BenchDeriveArgs) -> anyhow::Result<()> {
     let meili_client = Client::new(
         Some("http://127.0.0.1:7700".into()),
         args.master_key.as_deref(),
-        Some(std::time::Duration::from_secs(60)),
+        Some(std::time::Duration::from_secs(args.tasks_queue_timeout_secs)),
     )?;
 
     // enter runtime
@@ -135,7 +139,7 @@ pub fn run(args: BenchDeriveArgs) -> anyhow::Result<()> {
     rt.block_on(async {
         dashboard_client.send_machine_info(&env).await?;
 
-        let commit_message = build_info.commit_msg.context("missing commit message")?.split('\n').next().unwrap();
+        let commit_message = build_info.commit_msg.unwrap_or_default().split('\n').next().unwrap();
         let max_workloads = args.workload_file.len();
         let reason: Option<&str> = args.reason.as_deref();
         let invocation_uuid = dashboard_client.create_invocation(build_info.clone(), commit_message, env, max_workloads, reason).await?;

crates/xtask/src/main.rs

@@ -16,6 +16,7 @@ struct ListFeaturesDeriveArgs {
 #[command(author, version, about, long_about)]
 #[command(name = "cargo xtask")]
 #[command(bin_name = "cargo xtask")]
+#[allow(clippy::large_enum_variant)] // please, that's enough...
 enum Command {
     ListFeatures(ListFeaturesDeriveArgs),
     Bench(BenchDeriveArgs),

workloads/hackernews-add-new-documents.json

@@ -0,0 +1,105 @@
+{
+  "name": "hackernews.add_new_documents",
+  "run_count": 3,
+  "extra_cli_args": [],
+  "assets": {
+      "hackernews-01.ndjson": {
+        "local_location": null,
+        "remote_location": "https://milli-benchmarks.fra1.digitaloceanspaces.com/bench/datasets/hackernews/modification/01.ndjson",
+        "sha256": "cd3627b86c064d865b6754848ed0e73ef1d8142752a25e5f0765c3a1296dd3ae"
+      },
+      "hackernews-02.ndjson": {
+        "local_location": null,
+        "remote_location": "https://milli-benchmarks.fra1.digitaloceanspaces.com/bench/datasets/hackernews/modification/02.ndjson",
+        "sha256": "5d533b83bcf992201dace88b4d0c0be8b4df5225c6c4b763582d986844bcc23b"
+      },
+      "hackernews-03.ndjson": {
+        "local_location": null,
+        "remote_location": "https://milli-benchmarks.fra1.digitaloceanspaces.com/bench/datasets/hackernews/modification/03.ndjson",
+        "sha256": "f5f351a0d04a8a83643ace12cafa2b7ec8ca8cb7d46fd268e5126492a6c66f2a"
+      },
+      "hackernews-04.ndjson": {
+        "local_location": null,
+        "remote_location": "https://milli-benchmarks.fra1.digitaloceanspaces.com/bench/datasets/hackernews/modification/04.ndjson",
+        "sha256": "ac1915ee7ce53a6718548c255a6cc59969784b2570745dc5b739f714beda291a"
+      },
+      "hackernews-05.ndjson": {
+        "local_location": null,
+        "remote_location": "https://milli-benchmarks.fra1.digitaloceanspaces.com/bench/datasets/hackernews/modification/05.ndjson",
+        "sha256": "be31d5632602f798e62d1c10c83bdfda2b4deaa068477eacde05fdd247572b82"
+      }
+  },
+  "precommands": [
+    {
+      "route": "indexes/movies/settings",
+      "method": "PATCH",
+      "body": {
+        "inline": {
+          "displayedAttributes": [
+            "title",
+            "by",
+            "score",
+            "time",
+            "text"
+          ],
+          "searchableAttributes": [
+            "title",
+            "text"
+          ],
+          "filterableAttributes": [
+            "by",
+            "kids",
+            "parent"
+          ],
+          "sortableAttributes": [
+            "score",
+            "time"
+          ]
+        }
+      },
+      "synchronous": "WaitForTask"
+    },
+    {
+      "route": "indexes/movies/documents",
+      "method": "POST",
+      "body": {
+        "asset": "hackernews-01.ndjson"
+      },
+      "synchronous": "WaitForResponse"
+    },
+    {
+      "route": "indexes/movies/documents",
+      "method": "POST",
+      "body": {
+        "asset": "hackernews-02.ndjson"
+      },
+      "synchronous": "WaitForResponse"
+    },
+    {
+      "route": "indexes/movies/documents",
+      "method": "POST",
+      "body": {
+        "asset": "hackernews-03.ndjson"
+      },
+      "synchronous": "WaitForResponse"
+    },
+    {
+      "route": "indexes/movies/documents",
+      "method": "POST",
+      "body": {
+        "asset": "hackernews-04.ndjson"
+      },
+      "synchronous": "WaitForTask"
+    }
+  ],
+  "commands": [
+      {
+        "route": "indexes/movies/documents",
+        "method": "POST",
+        "body": {
+          "asset": "hackernews-05.ndjson"
+        },
+        "synchronous": "WaitForTask"
+      }
+  ]
+}

workloads/hackernews-modify-facet-numbers.json

@@ -0,0 +1,111 @@
+{
+  "name": "hackernews.modify_facet_numbers",
+  "run_count": 3,
+  "extra_cli_args": [],
+  "assets": {
+    "hackernews-01.ndjson": {
+      "local_location": null,
+      "remote_location": "https://milli-benchmarks.fra1.digitaloceanspaces.com/bench/datasets/hackernews/modification/01.ndjson",
+      "sha256": "cd3627b86c064d865b6754848ed0e73ef1d8142752a25e5f0765c3a1296dd3ae"
+    },
+    "hackernews-02.ndjson": {
+      "local_location": null,
+      "remote_location": "https://milli-benchmarks.fra1.digitaloceanspaces.com/bench/datasets/hackernews/modification/02.ndjson",
+      "sha256": "5d533b83bcf992201dace88b4d0c0be8b4df5225c6c4b763582d986844bcc23b"
+    },
+    "hackernews-03.ndjson": {
+      "local_location": null,
+      "remote_location": "https://milli-benchmarks.fra1.digitaloceanspaces.com/bench/datasets/hackernews/modification/03.ndjson",
+      "sha256": "f5f351a0d04a8a83643ace12cafa2b7ec8ca8cb7d46fd268e5126492a6c66f2a"
+    },
+    "hackernews-04.ndjson": {
+      "local_location": null,
+      "remote_location": "https://milli-benchmarks.fra1.digitaloceanspaces.com/bench/datasets/hackernews/modification/04.ndjson",
+      "sha256": "ac1915ee7ce53a6718548c255a6cc59969784b2570745dc5b739f714beda291a"
+    },
+    "hackernews-05.ndjson": {
+      "local_location": null,
+      "remote_location": "https://milli-benchmarks.fra1.digitaloceanspaces.com/bench/datasets/hackernews/modification/05.ndjson",
+      "sha256": "be31d5632602f798e62d1c10c83bdfda2b4deaa068477eacde05fdd247572b82"
+    },
+    "hackernews-02-modified-filters.ndjson": {
+      "local_location": null,
+      "remote_location": "https://milli-benchmarks.fra1.digitaloceanspaces.com/bench/datasets/hackernews/modification/02-modified-filters.ndjson",
+      "sha256": "7272cbfd41110d32d7fe168424a0000f07589bfe40f664652b34f4f20aaf3802"
+    }
+  },
+  "precommands": [
+    {
+      "route": "indexes/movies/settings",
+      "method": "PATCH",
+      "body": {
+        "inline": {
+          "displayedAttributes": [
+            "title",
+            "by",
+            "score",
+            "time",
+            "text"
+          ],
+          "searchableAttributes": [
+            "title",
+            "text"
+          ],
+          "filterableAttributes": [
+            "by",
+            "kids",
+            "parent"
+          ],
+          "sortableAttributes": [
+            "score",
+            "time"
+          ]
+        }
+      },
+      "synchronous": "WaitForTask"
+    },
+    {
+      "route": "indexes/movies/documents",
+      "method": "POST",
+      "body": {
+        "asset": "hackernews-01.ndjson"
+      },
+      "synchronous": "WaitForResponse"
+    },
+    {
+      "route": "indexes/movies/documents",
+      "method": "POST",
+      "body": {
+        "asset": "hackernews-02.ndjson"
+      },
+      "synchronous": "WaitForResponse"
+    },
+    {
+      "route": "indexes/movies/documents",
+      "method": "POST",
+      "body": {
+        "asset": "hackernews-03.ndjson"
+      },
+      "synchronous": "WaitForResponse"
+    },
+    {
+      "route": "indexes/movies/documents",
+      "method": "POST",
+      "body": {
+        "asset": "hackernews-04.ndjson"
+      },
+      "synchronous": "WaitForTask"
+    }
+  ],
+  "commands": [
+      {
+        "route": "indexes/movies/documents",
+        "method": "POST",
+        "body": {
+          "asset": "hackernews-02-modified-filters.ndjson"
+        },
+        "synchronous": "WaitForTask"
+      }
+  ]
+}
+  

workloads/hackernews-modify-facet-strings.json

@@ -0,0 +1,111 @@
+{
+  "name": "hackernews.modify_facet_strings",
+  "run_count": 3,
+  "extra_cli_args": [],
+  "assets": {
+    "hackernews-01.ndjson": {
+      "local_location": null,
+      "remote_location": "https://milli-benchmarks.fra1.digitaloceanspaces.com/bench/datasets/hackernews/modification/01.ndjson",
+      "sha256": "cd3627b86c064d865b6754848ed0e73ef1d8142752a25e5f0765c3a1296dd3ae"
+    },
+    "hackernews-02.ndjson": {
+      "local_location": null,
+      "remote_location": "https://milli-benchmarks.fra1.digitaloceanspaces.com/bench/datasets/hackernews/modification/02.ndjson",
+      "sha256": "5d533b83bcf992201dace88b4d0c0be8b4df5225c6c4b763582d986844bcc23b"
+    },
+    "hackernews-03.ndjson": {
+      "local_location": null,
+      "remote_location": "https://milli-benchmarks.fra1.digitaloceanspaces.com/bench/datasets/hackernews/modification/03.ndjson",
+      "sha256": "f5f351a0d04a8a83643ace12cafa2b7ec8ca8cb7d46fd268e5126492a6c66f2a"
+    },
+    "hackernews-04.ndjson": {
+      "local_location": null,
+      "remote_location": "https://milli-benchmarks.fra1.digitaloceanspaces.com/bench/datasets/hackernews/modification/04.ndjson",
+      "sha256": "ac1915ee7ce53a6718548c255a6cc59969784b2570745dc5b739f714beda291a"
+    },
+    "hackernews-05.ndjson": {
+      "local_location": null,
+      "remote_location": "https://milli-benchmarks.fra1.digitaloceanspaces.com/bench/datasets/hackernews/modification/05.ndjson",
+      "sha256": "be31d5632602f798e62d1c10c83bdfda2b4deaa068477eacde05fdd247572b82"
+    },
+    "hackernews-01-modified-filters.ndjson": {
+      "local_location": null,
+      "remote_location": "https://milli-benchmarks.fra1.digitaloceanspaces.com/bench/datasets/hackernews/modification/01-modified-filters.ndjson",
+      "sha256": "b80c245ce1b1df80b9b38800f677f3bd11947ebc62716fb108269d50e796c35c"
+    }
+  },
+  "precommands": [
+    {
+      "route": "indexes/movies/settings",
+      "method": "PATCH",
+      "body": {
+        "inline": {
+          "displayedAttributes": [
+            "title",
+            "by",
+            "score",
+            "time",
+            "text"
+          ],
+          "searchableAttributes": [
+            "title",
+            "text"
+          ],
+          "filterableAttributes": [
+            "by",
+            "kids",
+            "parent"
+          ],
+          "sortableAttributes": [
+            "score",
+            "time"
+          ]
+        }
+      },
+      "synchronous": "WaitForTask"
+    },
+    {
+      "route": "indexes/movies/documents",
+      "method": "POST",
+      "body": {
+        "asset": "hackernews-01.ndjson"
+      },
+      "synchronous": "WaitForResponse"
+    },
+    {
+      "route": "indexes/movies/documents",
+      "method": "POST",
+      "body": {
+        "asset": "hackernews-02.ndjson"
+      },
+      "synchronous": "WaitForResponse"
+    },
+    {
+      "route": "indexes/movies/documents",
+      "method": "POST",
+      "body": {
+        "asset": "hackernews-03.ndjson"
+      },
+      "synchronous": "WaitForResponse"
+    },
+    {
+      "route": "indexes/movies/documents",
+      "method": "POST",
+      "body": {
+        "asset": "hackernews-04.ndjson"
+      },
+      "synchronous": "WaitForTask"
+    }
+  ],
+  "commands": [
+      {
+        "route": "indexes/movies/documents",
+        "method": "POST",
+        "body": {
+          "asset": "hackernews-01-modified-filters.ndjson"
+        },
+        "synchronous": "WaitForTask"
+      }
+  ]
+}
+ 

workloads/hackernews-modify-searchables.json

@@ -0,0 +1,123 @@
+{
+  "name": "hackernews.modify_searchables",
+  "run_count": 3,
+  "extra_cli_args": [],
+  "assets": {
+    "hackernews-01.ndjson": {
+      "local_location": null,
+      "remote_location": "https://milli-benchmarks.fra1.digitaloceanspaces.com/bench/datasets/hackernews/modification/01.ndjson",
+      "sha256": "cd3627b86c064d865b6754848ed0e73ef1d8142752a25e5f0765c3a1296dd3ae"
+    },
+    "hackernews-02.ndjson": {
+      "local_location": null,
+      "remote_location": "https://milli-benchmarks.fra1.digitaloceanspaces.com/bench/datasets/hackernews/modification/02.ndjson",
+      "sha256": "5d533b83bcf992201dace88b4d0c0be8b4df5225c6c4b763582d986844bcc23b"
+    },
+    "hackernews-03.ndjson": {
+      "local_location": null,
+      "remote_location": "https://milli-benchmarks.fra1.digitaloceanspaces.com/bench/datasets/hackernews/modification/03.ndjson",
+      "sha256": "f5f351a0d04a8a83643ace12cafa2b7ec8ca8cb7d46fd268e5126492a6c66f2a"
+    },
+    "hackernews-04.ndjson": {
+      "local_location": null,
+      "remote_location": "https://milli-benchmarks.fra1.digitaloceanspaces.com/bench/datasets/hackernews/modification/04.ndjson",
+      "sha256": "ac1915ee7ce53a6718548c255a6cc59969784b2570745dc5b739f714beda291a"
+    },
+    "hackernews-05.ndjson": {
+      "local_location": null,
+      "remote_location": "https://milli-benchmarks.fra1.digitaloceanspaces.com/bench/datasets/hackernews/modification/05.ndjson",
+      "sha256": "be31d5632602f798e62d1c10c83bdfda2b4deaa068477eacde05fdd247572b82"
+    },
+    "hackernews-01-modified-searchables.ndjson": {
+      "local_location": null,
+      "remote_location": "https://milli-benchmarks.fra1.digitaloceanspaces.com/bench/datasets/hackernews/modification/01-modified-searchables.ndjson",
+      "sha256": "e5c08710c6af70031ac7212e0ba242c72ef29c8d4e1fce66c789544641452a7c"
+    },
+    "hackernews-02-modified-searchables.ndjson": {
+      "local_location": null,
+      "remote_location": "https://milli-benchmarks.fra1.digitaloceanspaces.com/bench/datasets/hackernews/modification/02-modified-searchables.ndjson",
+      "sha256": "098b029851117087b1e26ccb7ac408eda9bba54c3008213a2880d6fab607346e"
+    }
+  },
+  "precommands": [
+    {
+      "route": "indexes/movies/settings",
+      "method": "PATCH",
+      "body": {
+        "inline": {
+          "displayedAttributes": [
+            "title",
+            "by",
+            "score",
+            "time",
+            "text"
+          ],
+          "searchableAttributes": [
+            "title",
+            "text"
+          ],
+          "filterableAttributes": [
+            "by",
+            "kids",
+            "parent"
+          ],
+          "sortableAttributes": [
+            "score",
+            "time"
+          ]
+        }
+      },
+      "synchronous": "WaitForTask"
+    },
+    {
+      "route": "indexes/movies/documents",
+      "method": "POST",
+      "body": {
+        "asset": "hackernews-01.ndjson"
+      },
+      "synchronous": "WaitForResponse"
+    },
+    {
+      "route": "indexes/movies/documents",
+      "method": "POST",
+      "body": {
+        "asset": "hackernews-02.ndjson"
+      },
+      "synchronous": "WaitForResponse"
+    },
+    {
+      "route": "indexes/movies/documents",
+      "method": "POST",
+      "body": {
+        "asset": "hackernews-03.ndjson"
+      },
+      "synchronous": "WaitForResponse"
+    },
+    {
+      "route": "indexes/movies/documents",
+      "method": "POST",
+      "body": {
+        "asset": "hackernews-04.ndjson"
+      },
+      "synchronous": "WaitForTask"
+    }
+  ],
+  "commands": [
+      {
+        "route": "indexes/movies/documents",
+        "method": "POST",
+        "body": {
+          "asset": "hackernews-01-modified-searchables.ndjson"
+        },
+        "synchronous": "WaitForTask"
+      },
+      {
+        "route": "indexes/movies/documents",
+        "method": "POST",
+        "body": {
+          "asset": "hackernews-02-modified-searchables.ndjson"
+        },
+        "synchronous": "WaitForTask"
+      }
+  ]
+}