Compare commits

..

17 Commits

Author SHA1 Message Date
Clément Renault
5b7bdf3459 Only provide the last batch info 2025-11-20 12:03:11 +01:00
Kerollmops
6e5b42e924 remove-me: Introduce the MEILI_MAX_INDEX_SIZE env variable to limit the size of an index 2025-11-20 11:47:59 +01:00
Kerollmops
03ced42ad3 Fix some errors 2025-11-20 11:47:59 +01:00
Kerollmops
4706440610 Use a NonZeroUsize to store the prefix buffer length 2025-11-20 11:47:58 +01:00
Clément Renault
6700df1319 Fix the retrieval of the doc in scope 2025-11-20 11:47:58 +01:00
Kerollmops
a068931bd9 Allow users to delete documents 2025-11-20 11:47:58 +01:00
Clément Renault
ce87e14aa5 Erase a document if the AST is set 2025-11-20 11:47:58 +01:00
Clément Renault
7ff28985a0 Use the script in the edit documents 2025-11-20 11:47:58 +01:00
Clément Renault
a70aba8b68 Introduce a new executeAfterUpdate index setting 2025-11-20 11:47:58 +01:00
Clément Renault
2d8a61b53a Seems to work great, still need to read function from settings 2025-11-20 11:47:58 +01:00
Kerollmops
bd1fdc10e6 Expose the three last batches timings 2025-11-20 11:47:58 +01:00
Kerollmops
d1d6e219b3 Update utoipa 2025-11-20 11:47:58 +01:00
Kerollmops
b6a4f917f8 Reset metrics values to keep current steps only 2025-11-20 11:47:58 +01:00
Kerollmops
5c2a431d57 Expose the step currently running on the metrics route 2025-11-20 11:47:58 +01:00
Kerollmops
5781c16957 Simplify the auth filters 2025-11-20 11:47:58 +01:00
Kerollmops
246c44aeb7 Expose the metrics for the last finished batch and not the processing
one
2025-11-20 11:47:58 +01:00
Kerollmops
b2ed891f84 Expose batch progress traces on the metrics route 2025-11-20 11:47:57 +01:00
40 changed files with 370 additions and 184 deletions

41
Cargo.lock generated
View File

@@ -584,7 +584,7 @@ source = "git+https://github.com/meilisearch/bbqueue#cbb87cc707b5af415ef203bdaf2
[[package]]
name = "benchmarks"
version = "1.27.0"
version = "1.26.0"
dependencies = [
"anyhow",
"bumpalo",
@@ -794,7 +794,7 @@ dependencies = [
[[package]]
name = "build-info"
version = "1.27.0"
version = "1.26.0"
dependencies = [
"anyhow",
"time",
@@ -1784,7 +1784,7 @@ dependencies = [
[[package]]
name = "dump"
version = "1.27.0"
version = "1.26.0"
dependencies = [
"anyhow",
"big_s",
@@ -2027,7 +2027,7 @@ checksum = "37909eebbb50d72f9059c3b6d82c0463f2ff062c9e95845c43a6c9c0355411be"
[[package]]
name = "file-store"
version = "1.27.0"
version = "1.26.0"
dependencies = [
"tempfile",
"thiserror 2.0.16",
@@ -2049,7 +2049,7 @@ dependencies = [
[[package]]
name = "filter-parser"
version = "1.27.0"
version = "1.26.0"
dependencies = [
"insta",
"levenshtein_automata",
@@ -2077,7 +2077,7 @@ dependencies = [
[[package]]
name = "flatten-serde-json"
version = "1.27.0"
version = "1.26.0"
dependencies = [
"criterion",
"serde_json",
@@ -2234,7 +2234,7 @@ dependencies = [
[[package]]
name = "fuzzers"
version = "1.27.0"
version = "1.26.0"
dependencies = [
"arbitrary",
"bumpalo",
@@ -2922,6 +2922,12 @@ version = "1.0.3"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "df3b46402a9d5adb4c86a0cf463f42e19994e3ee891101b1841f30a545cb49a9"
[[package]]
name = "humantime"
version = "2.3.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "135b12329e5e3ce057a9f972339ea52bc954fe1e9358ef27f95e89716fbc5424"
[[package]]
name = "hyper"
version = "1.7.0"
@@ -3188,7 +3194,7 @@ dependencies = [
[[package]]
name = "index-scheduler"
version = "1.27.0"
version = "1.26.0"
dependencies = [
"anyhow",
"backoff",
@@ -3461,7 +3467,7 @@ dependencies = [
[[package]]
name = "json-depth-checker"
version = "1.27.0"
version = "1.26.0"
dependencies = [
"criterion",
"serde_json",
@@ -3980,7 +3986,7 @@ checksum = "490cc448043f947bae3cbee9c203358d62dbee0db12107a74be5c30ccfd09771"
[[package]]
name = "meili-snap"
version = "1.27.0"
version = "1.26.0"
dependencies = [
"insta",
"md5",
@@ -3991,7 +3997,7 @@ dependencies = [
[[package]]
name = "meilisearch"
version = "1.27.0"
version = "1.26.0"
dependencies = [
"actix-cors",
"actix-http",
@@ -4020,6 +4026,7 @@ dependencies = [
"futures",
"futures-util",
"hex",
"humantime",
"index-scheduler",
"indexmap",
"insta",
@@ -4088,7 +4095,7 @@ dependencies = [
[[package]]
name = "meilisearch-auth"
version = "1.27.0"
version = "1.26.0"
dependencies = [
"base64 0.22.1",
"enum-iterator",
@@ -4107,7 +4114,7 @@ dependencies = [
[[package]]
name = "meilisearch-types"
version = "1.27.0"
version = "1.26.0"
dependencies = [
"actix-web",
"anyhow",
@@ -4142,7 +4149,7 @@ dependencies = [
[[package]]
name = "meilitool"
version = "1.27.0"
version = "1.26.0"
dependencies = [
"anyhow",
"clap",
@@ -4176,7 +4183,7 @@ dependencies = [
[[package]]
name = "milli"
version = "1.27.0"
version = "1.26.0"
dependencies = [
"arroy",
"bbqueue",
@@ -4757,7 +4764,7 @@ checksum = "9b4f627cb1b25917193a259e49bdad08f671f8d9708acfd5fe0a8c1455d87220"
[[package]]
name = "permissive-json-pointer"
version = "1.27.0"
version = "1.26.0"
dependencies = [
"big_s",
"serde_json",
@@ -7879,7 +7886,7 @@ dependencies = [
[[package]]
name = "xtask"
version = "1.27.0"
version = "1.26.0"
dependencies = [
"anyhow",
"build-info",

View File

@@ -23,7 +23,7 @@ members = [
]
[workspace.package]
version = "1.27.0"
version = "1.26.0"
authors = [
"Quentin de Quelen <quentin@dequelen.me>",
"Clément Renault <clement@meilisearch.com>",

View File

@@ -341,6 +341,7 @@ pub(crate) mod test {
prefix_search: Setting::NotSet,
chat: Setting::NotSet,
vector_store: Setting::NotSet,
execute_after_update: Setting::NotSet,
_kind: std::marker::PhantomData,
};
settings.check()

View File

@@ -423,6 +423,7 @@ impl<T> From<v5::Settings<T>> for v6::Settings<v6::Unchecked> {
prefix_search: v6::Setting::NotSet,
chat: v6::Setting::NotSet,
vector_store: v6::Setting::NotSet,
execute_after_update: v6::Setting::NotSet,
_kind: std::marker::PhantomData,
}
}

View File

@@ -306,6 +306,18 @@ fn create_or_open_index(
) -> Result<Index> {
let options = EnvOpenOptions::new();
let mut options = options.read_txn_without_tls();
let map_size = match std::env::var("MEILI_MAX_INDEX_SIZE") {
Ok(max_size) => {
let max_size = max_size.parse().unwrap();
map_size.min(max_size)
}
Err(VarError::NotPresent) => map_size,
Err(VarError::NotUnicode(e)) => {
panic!("Non unicode max index size in `MEILI_MAX_INDEX_SIZE`: {e:?}")
}
};
options.map_size(clamp_to_page_size(map_size));
// You can find more details about this experimental

View File

@@ -6,7 +6,7 @@ source: crates/index-scheduler/src/scheduler/test_failure.rs
[]
----------------------------------------------------------------------
### All Tasks:
0 {uid: 0, batch_uid: 0, status: succeeded, details: { from: (1, 12, 0), to: (1, 27, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
0 {uid: 0, batch_uid: 0, status: succeeded, details: { from: (1, 12, 0), to: (1, 26, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
1 {uid: 1, batch_uid: 1, status: succeeded, details: { primary_key: Some("mouse"), old_new_uid: None, new_index_uid: None }, kind: IndexCreation { index_uid: "catto", primary_key: Some("mouse") }}
2 {uid: 2, batch_uid: 2, status: succeeded, details: { primary_key: Some("bone"), old_new_uid: None, new_index_uid: None }, kind: IndexCreation { index_uid: "doggo", primary_key: Some("bone") }}
3 {uid: 3, batch_uid: 3, status: failed, error: ResponseError { code: 200, message: "Index `doggo` already exists.", error_code: "index_already_exists", error_type: "invalid_request", error_link: "https://docs.meilisearch.com/errors#index_already_exists" }, details: { primary_key: Some("bone"), old_new_uid: None, new_index_uid: None }, kind: IndexCreation { index_uid: "doggo", primary_key: Some("bone") }}
@@ -57,7 +57,7 @@ girafo: { number_of_documents: 0, field_distribution: {} }
[timestamp] [4,]
----------------------------------------------------------------------
### All Batches:
0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.27.0"}, stats: {"totalNbTasks":1,"status":{"succeeded":1},"types":{"upgradeDatabase":1},"indexUids":{}}, stop reason: "stopped after the last task of type `upgradeDatabase` because they cannot be batched with tasks of any other type.", }
0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.26.0"}, stats: {"totalNbTasks":1,"status":{"succeeded":1},"types":{"upgradeDatabase":1},"indexUids":{}}, stop reason: "stopped after the last task of type `upgradeDatabase` because they cannot be batched with tasks of any other type.", }
1 {uid: 1, details: {"primaryKey":"mouse"}, stats: {"totalNbTasks":1,"status":{"succeeded":1},"types":{"indexCreation":1},"indexUids":{"catto":1}}, stop reason: "created batch containing only task with id 1 of type `indexCreation` that cannot be batched with any other task.", }
2 {uid: 2, details: {"primaryKey":"bone"}, stats: {"totalNbTasks":1,"status":{"succeeded":1},"types":{"indexCreation":1},"indexUids":{"doggo":1}}, stop reason: "created batch containing only task with id 2 of type `indexCreation` that cannot be batched with any other task.", }
3 {uid: 3, details: {"primaryKey":"bone"}, stats: {"totalNbTasks":1,"status":{"failed":1},"types":{"indexCreation":1},"indexUids":{"doggo":1}}, stop reason: "created batch containing only task with id 3 of type `indexCreation` that cannot be batched with any other task.", }

View File

@@ -6,7 +6,7 @@ source: crates/index-scheduler/src/scheduler/test_failure.rs
[]
----------------------------------------------------------------------
### All Tasks:
0 {uid: 0, status: enqueued, details: { from: (1, 12, 0), to: (1, 27, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
0 {uid: 0, status: enqueued, details: { from: (1, 12, 0), to: (1, 26, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
----------------------------------------------------------------------
### Status:
enqueued [0,]

View File

@@ -6,7 +6,7 @@ source: crates/index-scheduler/src/scheduler/test_failure.rs
[]
----------------------------------------------------------------------
### All Tasks:
0 {uid: 0, status: enqueued, details: { from: (1, 12, 0), to: (1, 27, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
0 {uid: 0, status: enqueued, details: { from: (1, 12, 0), to: (1, 26, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
1 {uid: 1, status: enqueued, details: { primary_key: Some("mouse"), old_new_uid: None, new_index_uid: None }, kind: IndexCreation { index_uid: "catto", primary_key: Some("mouse") }}
----------------------------------------------------------------------
### Status:

View File

@@ -6,7 +6,7 @@ source: crates/index-scheduler/src/scheduler/test_failure.rs
[]
----------------------------------------------------------------------
### All Tasks:
0 {uid: 0, batch_uid: 0, status: failed, error: ResponseError { code: 200, message: "Planned failure for tests.", error_code: "internal", error_type: "internal", error_link: "https://docs.meilisearch.com/errors#internal" }, details: { from: (1, 12, 0), to: (1, 27, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
0 {uid: 0, batch_uid: 0, status: failed, error: ResponseError { code: 200, message: "Planned failure for tests.", error_code: "internal", error_type: "internal", error_link: "https://docs.meilisearch.com/errors#internal" }, details: { from: (1, 12, 0), to: (1, 26, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
1 {uid: 1, status: enqueued, details: { primary_key: Some("mouse"), old_new_uid: None, new_index_uid: None }, kind: IndexCreation { index_uid: "catto", primary_key: Some("mouse") }}
----------------------------------------------------------------------
### Status:
@@ -37,7 +37,7 @@ catto [1,]
[timestamp] [0,]
----------------------------------------------------------------------
### All Batches:
0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.27.0"}, stats: {"totalNbTasks":1,"status":{"failed":1},"types":{"upgradeDatabase":1},"indexUids":{}}, stop reason: "stopped after the last task of type `upgradeDatabase` because they cannot be batched with tasks of any other type.", }
0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.26.0"}, stats: {"totalNbTasks":1,"status":{"failed":1},"types":{"upgradeDatabase":1},"indexUids":{}}, stop reason: "stopped after the last task of type `upgradeDatabase` because they cannot be batched with tasks of any other type.", }
----------------------------------------------------------------------
### Batch to tasks mapping:
0 [0,]

View File

@@ -6,7 +6,7 @@ source: crates/index-scheduler/src/scheduler/test_failure.rs
[]
----------------------------------------------------------------------
### All Tasks:
0 {uid: 0, batch_uid: 0, status: failed, error: ResponseError { code: 200, message: "Planned failure for tests.", error_code: "internal", error_type: "internal", error_link: "https://docs.meilisearch.com/errors#internal" }, details: { from: (1, 12, 0), to: (1, 27, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
0 {uid: 0, batch_uid: 0, status: failed, error: ResponseError { code: 200, message: "Planned failure for tests.", error_code: "internal", error_type: "internal", error_link: "https://docs.meilisearch.com/errors#internal" }, details: { from: (1, 12, 0), to: (1, 26, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
1 {uid: 1, status: enqueued, details: { primary_key: Some("mouse"), old_new_uid: None, new_index_uid: None }, kind: IndexCreation { index_uid: "catto", primary_key: Some("mouse") }}
2 {uid: 2, status: enqueued, details: { primary_key: Some("bone"), old_new_uid: None, new_index_uid: None }, kind: IndexCreation { index_uid: "doggo", primary_key: Some("bone") }}
----------------------------------------------------------------------
@@ -40,7 +40,7 @@ doggo [2,]
[timestamp] [0,]
----------------------------------------------------------------------
### All Batches:
0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.27.0"}, stats: {"totalNbTasks":1,"status":{"failed":1},"types":{"upgradeDatabase":1},"indexUids":{}}, stop reason: "stopped after the last task of type `upgradeDatabase` because they cannot be batched with tasks of any other type.", }
0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.26.0"}, stats: {"totalNbTasks":1,"status":{"failed":1},"types":{"upgradeDatabase":1},"indexUids":{}}, stop reason: "stopped after the last task of type `upgradeDatabase` because they cannot be batched with tasks of any other type.", }
----------------------------------------------------------------------
### Batch to tasks mapping:
0 [0,]

View File

@@ -6,7 +6,7 @@ source: crates/index-scheduler/src/scheduler/test_failure.rs
[]
----------------------------------------------------------------------
### All Tasks:
0 {uid: 0, batch_uid: 0, status: succeeded, details: { from: (1, 12, 0), to: (1, 27, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
0 {uid: 0, batch_uid: 0, status: succeeded, details: { from: (1, 12, 0), to: (1, 26, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
1 {uid: 1, status: enqueued, details: { primary_key: Some("mouse"), old_new_uid: None, new_index_uid: None }, kind: IndexCreation { index_uid: "catto", primary_key: Some("mouse") }}
2 {uid: 2, status: enqueued, details: { primary_key: Some("bone"), old_new_uid: None, new_index_uid: None }, kind: IndexCreation { index_uid: "doggo", primary_key: Some("bone") }}
3 {uid: 3, status: enqueued, details: { primary_key: Some("bone"), old_new_uid: None, new_index_uid: None }, kind: IndexCreation { index_uid: "doggo", primary_key: Some("bone") }}
@@ -43,7 +43,7 @@ doggo [2,3,]
[timestamp] [0,]
----------------------------------------------------------------------
### All Batches:
0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.27.0"}, stats: {"totalNbTasks":1,"status":{"succeeded":1},"types":{"upgradeDatabase":1},"indexUids":{}}, stop reason: "stopped after the last task of type `upgradeDatabase` because they cannot be batched with tasks of any other type.", }
0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.26.0"}, stats: {"totalNbTasks":1,"status":{"succeeded":1},"types":{"upgradeDatabase":1},"indexUids":{}}, stop reason: "stopped after the last task of type `upgradeDatabase` because they cannot be batched with tasks of any other type.", }
----------------------------------------------------------------------
### Batch to tasks mapping:
0 [0,]

View File

@@ -50,7 +50,6 @@ pub fn upgrade_index_scheduler(
(1, 24, _) => 0,
(1, 25, _) => 0,
(1, 26, _) => 0,
(1, 27, _) => 0,
(major, minor, patch) => {
if major > current_major
|| (major == current_major && minor > current_minor)

View File

@@ -324,6 +324,7 @@ InvalidSettingsDisplayedAttributes , InvalidRequest , BAD_REQU
InvalidSettingsDistinctAttribute , InvalidRequest , BAD_REQUEST ;
InvalidSettingsProximityPrecision , InvalidRequest , BAD_REQUEST ;
InvalidSettingsFacetSearch , InvalidRequest , BAD_REQUEST ;
InvalidSettingsexecuteAfterUpdate , InvalidRequest , BAD_REQUEST ;
InvalidSettingsPrefixSearch , InvalidRequest , BAD_REQUEST ;
InvalidSettingsFaceting , InvalidRequest , BAD_REQUEST ;
InvalidSettingsFilterableAttributes , InvalidRequest , BAD_REQUEST ;

View File

@@ -326,6 +326,12 @@ pub struct Settings<T> {
#[schema(value_type = Option<VectorStoreBackend>)]
pub vector_store: Setting<VectorStoreBackend>,
/// Function to execute after an update
#[serde(default, skip_serializing_if = "Setting::is_not_set")]
#[deserr(default, error = DeserrJsonError<InvalidSettingsexecuteAfterUpdate>)]
#[schema(value_type = Option<String>, example = json!("doc.likes += 1"))]
pub execute_after_update: Setting<String>,
#[serde(skip)]
#[deserr(skip)]
pub _kind: PhantomData<T>,
@@ -395,6 +401,7 @@ impl Settings<Checked> {
prefix_search: Setting::Reset,
chat: Setting::Reset,
vector_store: Setting::Reset,
execute_after_update: Setting::Reset,
_kind: PhantomData,
}
}
@@ -423,6 +430,7 @@ impl Settings<Checked> {
prefix_search,
chat,
vector_store,
execute_after_update,
_kind,
} = self;
@@ -449,6 +457,7 @@ impl Settings<Checked> {
prefix_search,
vector_store,
chat,
execute_after_update,
_kind: PhantomData,
}
}
@@ -501,6 +510,7 @@ impl Settings<Unchecked> {
prefix_search: self.prefix_search,
chat: self.chat,
vector_store: self.vector_store,
execute_after_update: self.execute_after_update,
_kind: PhantomData,
}
}
@@ -582,6 +592,10 @@ impl Settings<Unchecked> {
prefix_search: other.prefix_search.or(self.prefix_search),
chat: other.chat.clone().or(self.chat.clone()),
vector_store: other.vector_store.or(self.vector_store),
execute_after_update: other
.execute_after_update
.clone()
.or(self.execute_after_update.clone()),
_kind: PhantomData,
}
}
@@ -622,6 +636,7 @@ pub fn apply_settings_to_builder(
prefix_search,
chat,
vector_store,
execute_after_update,
_kind,
} = settings;
@@ -845,6 +860,14 @@ pub fn apply_settings_to_builder(
Setting::Reset => builder.reset_vector_store(),
Setting::NotSet => (),
}
match execute_after_update {
Setting::Set(execute_after_update) => {
builder.set_execute_after_update(execute_after_update.clone())
}
Setting::Reset => builder.reset_execute_after_update(),
Setting::NotSet => (),
}
}
pub enum SecretPolicy {
@@ -944,13 +967,13 @@ pub fn settings(
.collect();
let vector_store = index.get_vector_store(rtxn)?;
let embedders = Setting::Set(embedders);
let search_cutoff_ms = index.search_cutoff(rtxn)?;
let localized_attributes_rules = index.localized_attributes_rules(rtxn)?;
let prefix_search = index.prefix_search(rtxn)?.map(PrefixSearchSettings::from);
let facet_search = index.facet_search(rtxn)?;
let chat = index.chat_config(rtxn).map(ChatSettings::from)?;
let execute_after_update = index.execute_after_update(rtxn)?;
let mut settings = Settings {
displayed_attributes: match displayed_attributes {
@@ -995,6 +1018,10 @@ pub fn settings(
Some(vector_store) => Setting::Set(vector_store),
None => Setting::Reset,
},
execute_after_update: match execute_after_update {
Some(function) => Setting::Set(function.to_string()),
None => Setting::NotSet,
},
_kind: PhantomData,
};
@@ -1225,6 +1252,7 @@ pub(crate) mod test {
prefix_search: Setting::NotSet,
chat: Setting::NotSet,
vector_store: Setting::NotSet,
execute_after_update: Setting::NotSet,
_kind: PhantomData::<Unchecked>,
};
@@ -1258,7 +1286,7 @@ pub(crate) mod test {
prefix_search: Setting::NotSet,
chat: Setting::NotSet,
vector_store: Setting::NotSet,
execute_after_update: Setting::NotSet,
_kind: PhantomData::<Unchecked>,
};

View File

@@ -117,7 +117,7 @@ secrecy = "0.10.3"
actix-web-lab = { version = "0.24.1", default-features = false }
urlencoding = "2.1.3"
backoff = { version = "0.4.0", features = ["tokio"] }
humantime = { version = "2.3.0", default-features = false }
[dev-dependencies]
actix-rt = "2.10.0"

View File

@@ -231,14 +231,8 @@ pub fn setup_meilisearch(
max_number_of_tasks: 1_000_000,
max_number_of_batched_tasks: opt.experimental_max_number_of_batched_tasks,
batched_tasks_size_limit: opt.experimental_limit_batched_tasks_total_size.map_or_else(
|| {
opt.indexer_options
.max_indexing_memory
// By default, we use half of the available memory to determine the size of batched tasks
.map_or(u64::MAX, |mem| mem.as_u64() / 2)
// And never exceed 10 GiB when we infer the limit
.min(10 * 1024 * 1024 * 1024)
},
// By default, we use half of the available memory to determine the size of batched tasks
|| opt.indexer_options.max_indexing_memory.map_or(u64::MAX, |mem| mem.as_u64() / 2),
|size| size.as_u64(),
),
index_growth_amount: byte_unit::Byte::from_str("10GiB").unwrap().as_u64() as usize,

View File

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

View File

@@ -474,8 +474,7 @@ pub struct Opt {
pub experimental_max_number_of_batched_tasks: usize,
/// Experimentally controls the maximum total size, in bytes, of tasks that will be processed
/// simultaneously. When unspecified, defaults to half of the maximum indexing memory and
/// clamped to 10 GiB.
/// simultaneously. When unspecified, defaults to half of the maximum indexing memory.
///
/// See: <https://github.com/orgs/meilisearch/discussions/801>
#[clap(long, env = MEILI_EXPERIMENTAL_LIMIT_BATCHED_TASKS_TOTAL_SIZE)]

View File

@@ -498,6 +498,17 @@ make_setting_routes!(
camelcase_attr: "facetSearch",
analytics: FacetSearchAnalytics
},
{
route: "/execute-after-update",
update_verb: put,
value_type: String,
err_type: meilisearch_types::deserr::DeserrJsonError<
meilisearch_types::error::deserr_codes::InvalidSettingsexecuteAfterUpdate,
>,
attr: execute_after_update,
camelcase_attr: "executeAfterUpdate",
analytics: ExecuteAfterUpdateAnalytics
},
{
route: "/prefix-search",
update_verb: put,
@@ -619,6 +630,9 @@ pub async fn update_all(
new_settings.non_separator_tokens.as_ref().set(),
),
facet_search: FacetSearchAnalytics::new(new_settings.facet_search.as_ref().set()),
execute_after_update: ExecuteAfterUpdateAnalytics::new(
new_settings.execute_after_update.as_ref().set(),
),
prefix_search: PrefixSearchAnalytics::new(new_settings.prefix_search.as_ref().set()),
chat: ChatAnalytics::new(new_settings.chat.as_ref().set()),
vector_store: VectorStoreAnalytics::new(new_settings.vector_store.as_ref().set()),

View File

@@ -42,6 +42,7 @@ pub struct SettingsAnalytics {
pub prefix_search: PrefixSearchAnalytics,
pub chat: ChatAnalytics,
pub vector_store: VectorStoreAnalytics,
pub execute_after_update: ExecuteAfterUpdateAnalytics,
}
impl Aggregate for SettingsAnalytics {
@@ -197,6 +198,9 @@ impl Aggregate for SettingsAnalytics {
set: new.facet_search.set | self.facet_search.set,
value: new.facet_search.value.or(self.facet_search.value),
},
execute_after_update: ExecuteAfterUpdateAnalytics {
set: new.execute_after_update.set | self.execute_after_update.set,
},
prefix_search: PrefixSearchAnalytics {
set: new.prefix_search.set | self.prefix_search.set,
value: new.prefix_search.value.or(self.prefix_search.value),
@@ -669,6 +673,21 @@ impl FacetSearchAnalytics {
}
}
#[derive(Serialize, Default)]
pub struct ExecuteAfterUpdateAnalytics {
pub set: bool,
}
impl ExecuteAfterUpdateAnalytics {
pub fn new(distinct: Option<&String>) -> Self {
Self { set: distinct.is_some() }
}
pub fn into_settings(self) -> SettingsAnalytics {
SettingsAnalytics { execute_after_update: self, ..Default::default() }
}
}
#[derive(Serialize, Default)]
pub struct PrefixSearchAnalytics {
pub set: bool,

View File

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

View File

@@ -1339,117 +1339,3 @@ async fn get_document_with_vectors() {
}
"###);
}
#[actix_rt::test]
async fn test_fetch_documents_pagination_with_sorting() {
let server = Server::new_shared();
let index = server.unique_index();
let (task, _code) = index.create(None).await;
server.wait_task(task.uid()).await.succeeded();
// Set name as sortable attribute
let (task, code) = index.update_settings_sortable_attributes(json!(["name"])).await;
assert_eq!(code, 202);
server.wait_task(task.uid()).await.succeeded();
let documents = json!((0..50)
.map(|i| json!({"id": i, "name": format!("doc_{:05}", std::cmp::min(i, 5))}))
.collect::<Vec<_>>());
// Add documents as described in the bug report
let (task, code) = index.add_documents(documents, None).await;
assert_eq!(code, 202);
server.wait_task(task.uid()).await.succeeded();
// Request 1 (first page): offset 0, limit 2
let (response, code) = index
.fetch_documents(json!({
"offset": 0,
"limit": 2,
"sort": ["name:asc"]
}))
.await;
assert_eq!(code, 200);
let results = response["results"].as_array().unwrap();
snapshot!(json_string!(results), @r###"
[
{
"id": 0,
"name": "doc_00000"
},
{
"id": 1,
"name": "doc_00001"
}
]
"###);
// Request 2 (second page): offset 2, limit 2
let (response, code) = index
.fetch_documents(json!({
"offset": 2,
"limit": 2,
"sort": ["name:asc"]
}))
.await;
assert_eq!(code, 200);
let results = response["results"].as_array().unwrap();
snapshot!(json_string!(results), @r###"
[
{
"id": 2,
"name": "doc_00002"
},
{
"id": 3,
"name": "doc_00003"
}
]
"###);
// Request 3 (third page): offset 4, limit 2
let (response, code) = index
.fetch_documents(json!({
"offset": 4,
"limit": 2,
"sort": ["name:asc"]
}))
.await;
assert_eq!(code, 200);
let results = response["results"].as_array().unwrap();
snapshot!(json_string!(results), @r###"
[
{
"id": 4,
"name": "doc_00004"
},
{
"id": 5,
"name": "doc_00005"
}
]
"###);
// Request 4 (fourth page): offset 6, limit 2
let (response, code) = index
.fetch_documents(json!({
"offset": 6,
"limit": 2,
"sort": ["name:asc"]
}))
.await;
assert_eq!(code, 200);
let results = response["results"].as_array().unwrap();
snapshot!(json_string!(results), @r###"
[
{
"id": 6,
"name": "doc_00005"
},
{
"id": 7,
"name": "doc_00005"
}
]
"###);
}

View File

@@ -43,7 +43,7 @@ async fn version_too_old() {
std::fs::write(db_path.join("VERSION"), "1.11.9999").unwrap();
let options = Opt { experimental_dumpless_upgrade: true, ..default_settings };
let err = Server::new_with_options(options).await.map(|_| ()).unwrap_err();
snapshot!(err, @"Database version 1.11.9999 is too old for the experimental dumpless upgrade feature. Please generate a dump using the v1.11.9999 and import it in the v1.27.0");
snapshot!(err, @"Database version 1.11.9999 is too old for the experimental dumpless upgrade feature. Please generate a dump using the v1.11.9999 and import it in the v1.26.0");
}
#[actix_rt::test]
@@ -58,7 +58,7 @@ async fn version_requires_downgrade() {
std::fs::write(db_path.join("VERSION"), format!("{major}.{minor}.{patch}")).unwrap();
let options = Opt { experimental_dumpless_upgrade: true, ..default_settings };
let err = Server::new_with_options(options).await.map(|_| ()).unwrap_err();
snapshot!(err, @"Database version 1.27.1 is higher than the Meilisearch version 1.27.0. Downgrade is not supported");
snapshot!(err, @"Database version 1.26.1 is higher than the Meilisearch version 1.26.0. Downgrade is not supported");
}
#[actix_rt::test]

View File

@@ -8,7 +8,7 @@ source: crates/meilisearch/tests/upgrade/v1_12/v1_12_0.rs
"progress": null,
"details": {
"upgradeFrom": "v1.12.0",
"upgradeTo": "v1.27.0"
"upgradeTo": "v1.26.0"
},
"stats": {
"totalNbTasks": 1,

View File

@@ -8,7 +8,7 @@ source: crates/meilisearch/tests/upgrade/v1_12/v1_12_0.rs
"progress": null,
"details": {
"upgradeFrom": "v1.12.0",
"upgradeTo": "v1.27.0"
"upgradeTo": "v1.26.0"
},
"stats": {
"totalNbTasks": 1,

View File

@@ -8,7 +8,7 @@ source: crates/meilisearch/tests/upgrade/v1_12/v1_12_0.rs
"progress": null,
"details": {
"upgradeFrom": "v1.12.0",
"upgradeTo": "v1.27.0"
"upgradeTo": "v1.26.0"
},
"stats": {
"totalNbTasks": 1,

View File

@@ -12,7 +12,7 @@ source: crates/meilisearch/tests/upgrade/v1_12/v1_12_0.rs
"canceledBy": null,
"details": {
"upgradeFrom": "v1.12.0",
"upgradeTo": "v1.27.0"
"upgradeTo": "v1.26.0"
},
"error": null,
"duration": "[duration]",

View File

@@ -12,7 +12,7 @@ source: crates/meilisearch/tests/upgrade/v1_12/v1_12_0.rs
"canceledBy": null,
"details": {
"upgradeFrom": "v1.12.0",
"upgradeTo": "v1.27.0"
"upgradeTo": "v1.26.0"
},
"error": null,
"duration": "[duration]",

View File

@@ -12,7 +12,7 @@ source: crates/meilisearch/tests/upgrade/v1_12/v1_12_0.rs
"canceledBy": null,
"details": {
"upgradeFrom": "v1.12.0",
"upgradeTo": "v1.27.0"
"upgradeTo": "v1.26.0"
},
"error": null,
"duration": "[duration]",

View File

@@ -8,7 +8,7 @@ source: crates/meilisearch/tests/upgrade/v1_12/v1_12_0.rs
"progress": null,
"details": {
"upgradeFrom": "v1.12.0",
"upgradeTo": "v1.27.0"
"upgradeTo": "v1.26.0"
},
"stats": {
"totalNbTasks": 1,

View File

@@ -12,7 +12,7 @@ source: crates/meilisearch/tests/upgrade/v1_12/v1_12_0.rs
"canceledBy": null,
"details": {
"upgradeFrom": "v1.12.0",
"upgradeTo": "v1.27.0"
"upgradeTo": "v1.26.0"
},
"error": null,
"duration": "[duration]",

View File

@@ -87,7 +87,7 @@ impl Iterator for SortedDocumentsIterator<'_> {
};
// Otherwise don't directly iterate over children, skip them if we know we will go further
let mut to_skip = n;
let mut to_skip = n - 1;
while to_skip > 0 {
if let Err(e) = SortedDocumentsIterator::update_current(
current_child,
@@ -108,7 +108,7 @@ impl Iterator for SortedDocumentsIterator<'_> {
continue;
} else {
// The current iterator is large enough, so we can forward the call to it.
return inner.nth(to_skip);
return inner.nth(to_skip + 1);
}
}

View File

@@ -84,6 +84,7 @@ pub mod main_key {
pub const SEARCH_CUTOFF: &str = "search_cutoff";
pub const LOCALIZED_ATTRIBUTES_RULES: &str = "localized_attributes_rules";
pub const FACET_SEARCH: &str = "facet_search";
pub const EXECUTE_AFTER_UPDATE: &str = "execute-after-update";
pub const PREFIX_SEARCH: &str = "prefix_search";
pub const DOCUMENTS_STATS: &str = "documents_stats";
pub const DISABLED_TYPOS_TERMS: &str = "disabled_typos_terms";
@@ -1765,6 +1766,22 @@ impl Index {
self.main.remap_key_type::<Str>().delete(txn, main_key::CHAT)
}
pub fn execute_after_update<'t>(&self, txn: &'t RoTxn<'_>) -> heed::Result<Option<&'t str>> {
self.main.remap_types::<Str, Str>().get(txn, main_key::EXECUTE_AFTER_UPDATE)
}
pub(crate) fn put_execute_after_update(
&self,
txn: &mut RwTxn<'_>,
val: &str,
) -> heed::Result<()> {
self.main.remap_types::<Str, Str>().put(txn, main_key::EXECUTE_AFTER_UPDATE, &val)
}
pub(crate) fn delete_execute_after_update(&self, txn: &mut RwTxn<'_>) -> heed::Result<bool> {
self.main.remap_key_type::<Str>().delete(txn, main_key::EXECUTE_AFTER_UPDATE)
}
pub fn localized_attributes_rules(
&self,
rtxn: &RoTxn<'_>,

View File

@@ -470,6 +470,71 @@ impl<'doc> Versions<'doc> {
Ok(Some(Self::single(data)))
}
pub fn multiple_with_edits(
doc: Option<rhai::Map>,
mut versions: impl Iterator<Item = Result<RawMap<'doc, FxBuildHasher>>>,
engine: &rhai::Engine,
edit_function: &rhai::AST,
doc_alloc: &'doc bumpalo::Bump,
) -> Result<Option<Option<Self>>> {
let Some(data) = versions.next() else { return Ok(None) };
let mut doc = doc.unwrap_or_default();
let mut data = data?;
for version in versions {
let version = version?;
for (field, value) in version {
data.insert(field, value);
}
let mut scope = rhai::Scope::new();
data.iter().for_each(|(k, v)| {
doc.insert(k.into(), serde_json::from_str(v.get()).unwrap());
});
scope.push("doc", doc.clone());
let _ = engine.eval_ast_with_scope::<rhai::Dynamic>(&mut scope, edit_function).unwrap();
data = RawMap::with_hasher_in(FxBuildHasher, doc_alloc);
match scope.get_value::<rhai::Map>("doc") {
Some(map) => {
for (key, value) in map {
let mut vec = bumpalo::collections::Vec::new_in(doc_alloc);
serde_json::to_writer(&mut vec, &value).unwrap();
let key = doc_alloc.alloc_str(key.as_str());
let raw_value = serde_json::from_slice(vec.into_bump_slice()).unwrap();
data.insert(key, raw_value);
}
}
// In case the deletes the document and it's not the last change
// we simply set the document to an empty one and await the next change.
None => (),
}
}
// We must also run the code after the last change
let mut scope = rhai::Scope::new();
data.iter().for_each(|(k, v)| {
doc.insert(k.into(), serde_json::from_str(v.get()).unwrap());
});
scope.push("doc", doc);
let _ = engine.eval_ast_with_scope::<rhai::Dynamic>(&mut scope, edit_function).unwrap();
data = RawMap::with_hasher_in(FxBuildHasher, doc_alloc);
match scope.get_value::<rhai::Map>("doc") {
Some(map) => {
for (key, value) in map {
let mut vec = bumpalo::collections::Vec::new_in(doc_alloc);
serde_json::to_writer(&mut vec, &value).unwrap();
let key = doc_alloc.alloc_str(key.as_str());
let raw_value = serde_json::from_slice(vec.into_bump_slice()).unwrap();
data.insert(key, raw_value);
}
Ok(Some(Some(Self::single(data))))
}
None => Ok(Some(None)),
}
}
pub fn single(version: RawMap<'doc, FxBuildHasher>) -> Self {
Self { data: version }
}

View File

@@ -18,6 +18,7 @@ use crate::documents::PrimaryKey;
use crate::progress::{AtomicPayloadStep, Progress};
use crate::update::new::document::{DocumentContext, Versions};
use crate::update::new::indexer::enterprise_edition::sharding::Shards;
use crate::update::new::indexer::update_by_function::obkv_to_rhaimap;
use crate::update::new::steps::IndexingStep;
use crate::update::new::thread_local::MostlySend;
use crate::update::new::{DocumentIdentifiers, Insertion, Update};
@@ -162,7 +163,16 @@ impl<'pl> DocumentOperation<'pl> {
.sort_unstable_by_key(|(_, po)| first_update_pointer(&po.operations).unwrap_or(0));
let docids_version_offsets = docids_version_offsets.into_bump_slice();
Ok((DocumentOperationChanges { docids_version_offsets }, operations_stats, primary_key))
let engine = rhai::Engine::new();
// Make sure to correctly setup the engine and remove all settings
let ast = index.execute_after_update(rtxn)?.map(|f| engine.compile(f).unwrap());
let fidmap = index.fields_ids_map(rtxn)?;
Ok((
DocumentOperationChanges { docids_version_offsets, engine, ast, fidmap },
operations_stats,
primary_key,
))
}
}
@@ -435,7 +445,15 @@ impl<'pl> DocumentChanges<'pl> for DocumentOperationChanges<'pl> {
'pl: 'doc,
{
let (external_doc, payload_operations) = item;
payload_operations.merge(external_doc, &context.doc_alloc)
payload_operations.merge(
&context.rtxn,
context.index,
&self.fidmap,
&self.engine,
self.ast.as_ref(),
external_doc,
&context.doc_alloc,
)
}
fn len(&self) -> usize {
@@ -444,6 +462,9 @@ impl<'pl> DocumentChanges<'pl> for DocumentOperationChanges<'pl> {
}
pub struct DocumentOperationChanges<'pl> {
engine: rhai::Engine,
ast: Option<rhai::AST>,
fidmap: FieldsIdsMap,
docids_version_offsets: &'pl [(&'pl str, PayloadOperations<'pl>)],
}
@@ -506,10 +527,14 @@ impl<'pl> PayloadOperations<'pl> {
}
/// Returns only the most recent version of a document based on the updates from the payloads.
///
/// This function is only meant to be used when doing a replacement and not an update.
#[allow(clippy::too_many_arguments)]
fn merge<'doc>(
&self,
rtxn: &heed::RoTxn,
index: &Index,
fidmap: &FieldsIdsMap,
engine: &rhai::Engine,
ast: Option<&rhai::AST>,
external_doc: &'doc str,
doc_alloc: &'doc Bump,
) -> Result<Option<DocumentChange<'doc>>>
@@ -573,9 +598,33 @@ impl<'pl> PayloadOperations<'pl> {
Ok(document)
});
let Some(versions) = Versions::multiple(versions)? else { return Ok(None) };
let versions = match ast {
Some(ast) => {
let doc = index
.documents
.get(rtxn, &self.docid)?
.map(|obkv| obkv_to_rhaimap(obkv, fidmap))
.transpose()?;
match Versions::multiple_with_edits(doc, versions, engine, ast, doc_alloc)?
{
Some(Some(versions)) => Some(versions),
Some(None) if self.is_new => return Ok(None),
Some(None) => {
return Ok(Some(DocumentChange::Deletion(
DocumentIdentifiers::create(self.docid, external_doc),
)));
}
None => None,
}
}
None => Versions::multiple(versions)?,
};
if self.is_new {
let Some(versions) = versions else {
return Ok(None);
};
if self.is_new || ast.is_some() {
Ok(Some(DocumentChange::Insertion(Insertion::create(
self.docid,
external_doc,

View File

@@ -187,7 +187,7 @@ impl<'index> DocumentChanges<'index> for UpdateByFunctionChanges<'index> {
}
}
fn obkv_to_rhaimap(obkv: &KvReaderFieldId, fields_ids_map: &FieldsIdsMap) -> Result<rhai::Map> {
pub fn obkv_to_rhaimap(obkv: &KvReaderFieldId, fields_ids_map: &FieldsIdsMap) -> Result<rhai::Map> {
let all_keys = obkv.iter().map(|(k, _v)| k).collect::<Vec<_>>();
let map: Result<rhai::Map> = all_keys
.iter()

View File

@@ -2,6 +2,7 @@ use std::cell::RefCell;
use std::collections::BTreeSet;
use std::io::{BufReader, BufWriter, Read, Seek, Write};
use std::iter;
use std::num::NonZeroUsize;
use hashbrown::HashMap;
use heed::types::{Bytes, DecodeIgnore, Str};
@@ -345,7 +346,7 @@ impl<'i> WordPrefixIntegerDocids<'i> {
indexes.push(PrefixIntegerEntry {
prefix,
pos,
serialized_length: Some(buffer.len()),
serialized_length: NonZeroUsize::new(buffer.len()),
});
file.write_all(&buffer)?;
}
@@ -371,7 +372,7 @@ impl<'i> WordPrefixIntegerDocids<'i> {
key_buffer.extend_from_slice(&pos.to_be_bytes());
match serialized_length {
Some(serialized_length) => {
buffer.resize(serialized_length, 0);
buffer.resize(serialized_length.get(), 0);
file.read_exact(&mut buffer)?;
self.prefix_database.remap_data_type::<Bytes>().put(
wtxn,
@@ -426,7 +427,7 @@ impl<'i> WordPrefixIntegerDocids<'i> {
index.push(PrefixIntegerEntry {
prefix,
pos,
serialized_length: Some(buffer.len()),
serialized_length: NonZeroUsize::new(buffer.len()),
});
file.write_all(buffer)?;
}
@@ -452,7 +453,7 @@ impl<'i> WordPrefixIntegerDocids<'i> {
key_buffer.extend_from_slice(&pos.to_be_bytes());
match serialized_length {
Some(serialized_length) => {
buffer.resize(serialized_length, 0);
buffer.resize(serialized_length.get(), 0);
file.read_exact(&mut buffer)?;
self.prefix_database.remap_data_type::<Bytes>().put(
wtxn,
@@ -475,7 +476,7 @@ impl<'i> WordPrefixIntegerDocids<'i> {
struct PrefixIntegerEntry<'a> {
prefix: &'a str,
pos: u16,
serialized_length: Option<usize>,
serialized_length: Option<NonZeroUsize>,
}
/// TODO doc

View File

@@ -202,6 +202,7 @@ pub struct Settings<'a, 't, 'i> {
facet_search: Setting<bool>,
chat: Setting<ChatSettings>,
vector_store: Setting<VectorStoreBackend>,
execute_after_update: Setting<String>,
}
impl<'a, 't, 'i> Settings<'a, 't, 'i> {
@@ -242,6 +243,7 @@ impl<'a, 't, 'i> Settings<'a, 't, 'i> {
facet_search: Setting::NotSet,
chat: Setting::NotSet,
vector_store: Setting::NotSet,
execute_after_update: Setting::NotSet,
indexer_config,
}
}
@@ -488,6 +490,14 @@ impl<'a, 't, 'i> Settings<'a, 't, 'i> {
self.vector_store = Setting::Reset;
}
pub fn set_execute_after_update(&mut self, value: String) {
self.execute_after_update = Setting::Set(value);
}
pub fn reset_execute_after_update(&mut self) {
self.execute_after_update = Setting::Reset;
}
#[tracing::instrument(
level = "trace"
skip(self, progress_callback, should_abort, settings_diff, embedder_stats),
@@ -1059,6 +1069,18 @@ impl<'a, 't, 'i> Settings<'a, 't, 'i> {
Ok(changed)
}
fn update_execute_after_update(&mut self) -> Result<()> {
match self.execute_after_update.as_ref() {
Setting::Set(new) => {
self.index.put_execute_after_update(self.wtxn, &new).map_err(Into::into)
}
Setting::Reset => {
self.index.delete_execute_after_update(self.wtxn).map(drop).map_err(Into::into)
}
Setting::NotSet => Ok(()),
}
}
fn update_embedding_configs(&mut self) -> Result<BTreeMap<String, EmbedderAction>> {
match std::mem::take(&mut self.embedder_settings) {
Setting::Set(configs) => self.update_embedding_configs_set(configs),
@@ -1469,6 +1491,7 @@ impl<'a, 't, 'i> Settings<'a, 't, 'i> {
self.update_proximity_precision()?;
self.update_prefix_search()?;
self.update_facet_search()?;
self.update_execute_after_update()?;
self.update_localized_attributes_rules()?;
self.update_disabled_typos_terms()?;
self.update_chat_config()?;
@@ -1618,6 +1641,7 @@ impl<'a, 't, 'i> Settings<'a, 't, 'i> {
disable_on_numbers: Setting::NotSet,
chat: Setting::NotSet,
vector_store: Setting::NotSet,
execute_after_update: Setting::NotSet,
wtxn: _,
index: _,
indexer_config: _,

View File

@@ -899,6 +899,7 @@ fn test_correct_settings_init() {
disable_on_numbers,
chat,
vector_store,
execute_after_update,
} = settings;
assert!(matches!(searchable_fields, Setting::NotSet));
assert!(matches!(displayed_fields, Setting::NotSet));
@@ -929,6 +930,7 @@ fn test_correct_settings_init() {
assert!(matches!(disable_on_numbers, Setting::NotSet));
assert!(matches!(chat, Setting::NotSet));
assert!(matches!(vector_store, Setting::NotSet));
assert!(matches!(execute_after_update, Setting::NotSet));
})
.unwrap();
}

View File

@@ -44,7 +44,6 @@ const UPGRADE_FUNCTIONS: &[&dyn UpgradeIndex] = &[
&ToTargetNoOp { target: (1, 24, 0) },
&ToTargetNoOp { target: (1, 25, 0) },
&ToTargetNoOp { target: (1, 26, 0) },
&ToTargetNoOp { target: (1, 27, 0) },
// This is the last upgrade function, it will be called when the index is up to date.
// any other upgrade function should be added before this one.
&ToCurrentNoOp {},
@@ -82,7 +81,6 @@ const fn start(from: (u32, u32, u32)) -> Option<usize> {
(1, 24, _) => function_index!(14),
(1, 25, _) => function_index!(15),
(1, 26, _) => function_index!(16),
(1, 27, _) => function_index!(17),
// We deliberately don't add a placeholder with (VERSION_MAJOR, VERSION_MINOR, VERSION_PATCH) here to force manually
// considering dumpless upgrade.
(_major, _minor, _patch) => return None,