mirror of
https://github.com/meilisearch/MeiliSearch
synced 2024-12-04 18:45:46 +01:00
Add the options into the IndexScheduler
This commit is contained in:
parent
966cbdab69
commit
76657af1f9
@ -135,7 +135,7 @@ impl IndexMapper {
|
|||||||
index_growth_amount: usize,
|
index_growth_amount: usize,
|
||||||
index_count: usize,
|
index_count: usize,
|
||||||
enable_mdb_writemap: bool,
|
enable_mdb_writemap: bool,
|
||||||
indexer_config: IndexerConfig,
|
indexer_config: Arc<IndexerConfig>,
|
||||||
) -> Result<Self> {
|
) -> Result<Self> {
|
||||||
let mut wtxn = env.write_txn()?;
|
let mut wtxn = env.write_txn()?;
|
||||||
let index_mapping = env.create_database(&mut wtxn, Some(INDEX_MAPPING))?;
|
let index_mapping = env.create_database(&mut wtxn, Some(INDEX_MAPPING))?;
|
||||||
@ -150,7 +150,7 @@ impl IndexMapper {
|
|||||||
index_base_map_size,
|
index_base_map_size,
|
||||||
index_growth_amount,
|
index_growth_amount,
|
||||||
enable_mdb_writemap,
|
enable_mdb_writemap,
|
||||||
indexer_config: Arc::new(indexer_config),
|
indexer_config,
|
||||||
})
|
})
|
||||||
}
|
}
|
||||||
|
|
||||||
|
@ -265,7 +265,7 @@ pub struct IndexSchedulerOptions {
|
|||||||
/// The number of indexes that can be concurrently opened in memory.
|
/// The number of indexes that can be concurrently opened in memory.
|
||||||
pub index_count: usize,
|
pub index_count: usize,
|
||||||
/// Configuration used during indexing for each meilisearch index.
|
/// Configuration used during indexing for each meilisearch index.
|
||||||
pub indexer_config: IndexerConfig,
|
pub indexer_config: Arc<IndexerConfig>,
|
||||||
/// Set to `true` iff the index scheduler is allowed to automatically
|
/// Set to `true` iff the index scheduler is allowed to automatically
|
||||||
/// batch tasks together, to process multiple tasks at once.
|
/// batch tasks together, to process multiple tasks at once.
|
||||||
pub autobatching_enabled: bool,
|
pub autobatching_enabled: bool,
|
||||||
@ -290,7 +290,7 @@ pub struct IndexScheduler {
|
|||||||
impl IndexScheduler {
|
impl IndexScheduler {
|
||||||
/// Create an index scheduler and start its run loop.
|
/// Create an index scheduler and start its run loop.
|
||||||
pub fn new(
|
pub fn new(
|
||||||
options: IndexSchedulerOptions,
|
options: Arc<IndexSchedulerOptions>,
|
||||||
#[cfg(test)] test_breakpoint_sdr: crossbeam::channel::Sender<(Breakpoint, bool)>,
|
#[cfg(test)] test_breakpoint_sdr: crossbeam::channel::Sender<(Breakpoint, bool)>,
|
||||||
#[cfg(test)] planned_failures: Vec<(usize, tests::FailureLocation)>,
|
#[cfg(test)] planned_failures: Vec<(usize, tests::FailureLocation)>,
|
||||||
) -> Result<Self> {
|
) -> Result<Self> {
|
||||||
@ -898,6 +898,9 @@ pub struct IndexSchedulerInner {
|
|||||||
/// The LMDB environment which the DBs are associated with.
|
/// The LMDB environment which the DBs are associated with.
|
||||||
pub(crate) env: Env,
|
pub(crate) env: Env,
|
||||||
|
|
||||||
|
/// The options to open an IndexScheduler.
|
||||||
|
pub(crate) options: Arc<IndexSchedulerOptions>,
|
||||||
|
|
||||||
/// A boolean that can be set to true to stop the currently processing tasks.
|
/// A boolean that can be set to true to stop the currently processing tasks.
|
||||||
pub(crate) must_stop_processing: MustStopProcessing,
|
pub(crate) must_stop_processing: MustStopProcessing,
|
||||||
|
|
||||||
@ -982,7 +985,7 @@ pub struct IndexSchedulerInner {
|
|||||||
|
|
||||||
impl IndexSchedulerInner {
|
impl IndexSchedulerInner {
|
||||||
fn new(
|
fn new(
|
||||||
options: IndexSchedulerOptions,
|
options: Arc<IndexSchedulerOptions>,
|
||||||
#[cfg(test)] test_breakpoint_sdr: crossbeam::channel::Sender<(Breakpoint, bool)>,
|
#[cfg(test)] test_breakpoint_sdr: crossbeam::channel::Sender<(Breakpoint, bool)>,
|
||||||
#[cfg(test)] planned_failures: Vec<(usize, tests::FailureLocation)>,
|
#[cfg(test)] planned_failures: Vec<(usize, tests::FailureLocation)>,
|
||||||
) -> Result<Self> {
|
) -> Result<Self> {
|
||||||
@ -1015,7 +1018,7 @@ impl IndexSchedulerInner {
|
|||||||
let env = heed::EnvOpenOptions::new()
|
let env = heed::EnvOpenOptions::new()
|
||||||
.max_dbs(11)
|
.max_dbs(11)
|
||||||
.map_size(budget.task_db_size)
|
.map_size(budget.task_db_size)
|
||||||
.open(options.tasks_path)?;
|
.open(&options.tasks_path)?;
|
||||||
|
|
||||||
let features = features::FeatureData::new(&env, options.instance_features)?;
|
let features = features::FeatureData::new(&env, options.instance_features)?;
|
||||||
|
|
||||||
@ -1047,23 +1050,24 @@ impl IndexSchedulerInner {
|
|||||||
finished_at,
|
finished_at,
|
||||||
index_mapper: IndexMapper::new(
|
index_mapper: IndexMapper::new(
|
||||||
&env,
|
&env,
|
||||||
options.indexes_path,
|
options.indexes_path.clone(),
|
||||||
budget.map_size,
|
budget.map_size,
|
||||||
options.index_growth_amount,
|
options.index_growth_amount,
|
||||||
budget.index_count,
|
budget.index_count,
|
||||||
options.enable_mdb_writemap,
|
options.enable_mdb_writemap,
|
||||||
options.indexer_config,
|
options.indexer_config.clone(),
|
||||||
)?,
|
)?,
|
||||||
env,
|
env,
|
||||||
// we want to start the loop right away in case meilisearch was ctrl+Ced while processing things
|
// we want to start the loop right away in case meilisearch was ctrl+Ced while processing things
|
||||||
wake_up: Arc::new(SignalEvent::auto(true)),
|
wake_up: Arc::new(SignalEvent::auto(true)),
|
||||||
autobatching_enabled: options.autobatching_enabled,
|
autobatching_enabled: options.autobatching_enabled,
|
||||||
max_number_of_tasks: options.max_number_of_tasks,
|
max_number_of_tasks: options.max_number_of_tasks,
|
||||||
dumps_path: options.dumps_path,
|
dumps_path: options.dumps_path.clone(),
|
||||||
snapshots_path: options.snapshots_path,
|
snapshots_path: options.snapshots_path.clone(),
|
||||||
auth_path: options.auth_path,
|
auth_path: options.auth_path.clone(),
|
||||||
version_file_path: options.version_file_path,
|
version_file_path: options.version_file_path.clone(),
|
||||||
zookeeper: options.zookeeper,
|
zookeeper: options.zookeeper.clone(),
|
||||||
|
options,
|
||||||
#[cfg(test)]
|
#[cfg(test)]
|
||||||
test_breakpoint_sdr,
|
test_breakpoint_sdr,
|
||||||
#[cfg(test)]
|
#[cfg(test)]
|
||||||
|
@ -224,7 +224,7 @@ fn open_or_create_database_unchecked(
|
|||||||
// wrap our two builders in a closure that'll be executed later.
|
// wrap our two builders in a closure that'll be executed later.
|
||||||
let auth_controller = AuthController::new(&opt.db_path, &opt.master_key, zookeeper.clone());
|
let auth_controller = AuthController::new(&opt.db_path, &opt.master_key, zookeeper.clone());
|
||||||
let instance_features = opt.to_instance_features();
|
let instance_features = opt.to_instance_features();
|
||||||
let index_scheduler = IndexScheduler::new(IndexSchedulerOptions {
|
let index_scheduler = IndexScheduler::new(Arc::new(IndexSchedulerOptions {
|
||||||
version_file_path: opt.db_path.join(VERSION_FILE_NAME),
|
version_file_path: opt.db_path.join(VERSION_FILE_NAME),
|
||||||
auth_path: opt.db_path.join("auth"),
|
auth_path: opt.db_path.join("auth"),
|
||||||
tasks_path: opt.db_path.join("tasks"),
|
tasks_path: opt.db_path.join("tasks"),
|
||||||
@ -235,14 +235,14 @@ fn open_or_create_database_unchecked(
|
|||||||
task_db_size: opt.max_task_db_size.get_bytes() as usize,
|
task_db_size: opt.max_task_db_size.get_bytes() as usize,
|
||||||
index_base_map_size: opt.max_index_size.get_bytes() as usize,
|
index_base_map_size: opt.max_index_size.get_bytes() as usize,
|
||||||
enable_mdb_writemap: opt.experimental_reduce_indexing_memory_usage,
|
enable_mdb_writemap: opt.experimental_reduce_indexing_memory_usage,
|
||||||
indexer_config: (&opt.indexer_options).try_into()?,
|
indexer_config: (&opt.indexer_options).try_into().map(Arc::new)?,
|
||||||
autobatching_enabled: true,
|
autobatching_enabled: true,
|
||||||
max_number_of_tasks: 1_000_000,
|
max_number_of_tasks: 1_000_000,
|
||||||
index_growth_amount: byte_unit::Byte::from_str("10GiB").unwrap().get_bytes() as usize,
|
index_growth_amount: byte_unit::Byte::from_str("10GiB").unwrap().get_bytes() as usize,
|
||||||
index_count: DEFAULT_INDEX_COUNT,
|
index_count: DEFAULT_INDEX_COUNT,
|
||||||
instance_features,
|
instance_features,
|
||||||
zookeeper: zookeeper.clone(),
|
zookeeper: zookeeper.clone(),
|
||||||
})
|
}))
|
||||||
.map_err(anyhow::Error::from);
|
.map_err(anyhow::Error::from);
|
||||||
|
|
||||||
match (
|
match (
|
||||||
|
@ -557,7 +557,7 @@ impl TryFrom<&IndexerOpts> for IndexerConfig {
|
|||||||
Ok(Self {
|
Ok(Self {
|
||||||
log_every_n: Some(DEFAULT_LOG_EVERY_N),
|
log_every_n: Some(DEFAULT_LOG_EVERY_N),
|
||||||
max_memory: other.max_indexing_memory.map(|b| b.get_bytes() as usize),
|
max_memory: other.max_indexing_memory.map(|b| b.get_bytes() as usize),
|
||||||
thread_pool: Some(thread_pool),
|
thread_pool: Some(Arc::new(thread_pool)),
|
||||||
max_positions_per_attributes: None,
|
max_positions_per_attributes: None,
|
||||||
skip_index_budget: other.skip_index_budget,
|
skip_index_budget: other.skip_index_budget,
|
||||||
..Default::default()
|
..Default::default()
|
||||||
|
@ -1,3 +1,5 @@
|
|||||||
|
use std::sync::Arc;
|
||||||
|
|
||||||
use grenad::CompressionType;
|
use grenad::CompressionType;
|
||||||
use rayon::ThreadPool;
|
use rayon::ThreadPool;
|
||||||
|
|
||||||
@ -9,7 +11,7 @@ pub struct IndexerConfig {
|
|||||||
pub max_memory: Option<usize>,
|
pub max_memory: Option<usize>,
|
||||||
pub chunk_compression_type: CompressionType,
|
pub chunk_compression_type: CompressionType,
|
||||||
pub chunk_compression_level: Option<u32>,
|
pub chunk_compression_level: Option<u32>,
|
||||||
pub thread_pool: Option<ThreadPool>,
|
pub thread_pool: Option<Arc<ThreadPool>>,
|
||||||
pub max_positions_per_attributes: Option<u32>,
|
pub max_positions_per_attributes: Option<u32>,
|
||||||
pub skip_index_budget: bool,
|
pub skip_index_budget: bool,
|
||||||
}
|
}
|
||||||
|
Loading…
Reference in New Issue
Block a user