snowykami/meilisearch
1
0
Fork 0
mirror of https://github.com/meilisearch/meilisearch.git synced 2024-11-22 10:07:40 +08:00
Code Issues Actions 9 Packages Projects Releases Wiki Activity

Build mergers in parallel

Browse Source
This commit is contained in:
ManyTheFish 2024-09-11 11:49:26 +02:00
parent 39b5990f64
commit 2b317c681b
3 changed files with 95 additions and 53 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

57
milli/src/update/new/extract/searchable/extract_word_docids.rs
View File

@ -426,21 +426,38 @@ impl WordDocidsMergerBuilders {
current_docid: _,
} = other;
let sorter = word_fid_docids.into_sorter()?;
let readers = sorter.into_reader_cursors()?;
self.word_fid_docids.extend(readers);
let sorter = word_docids.into_sorter()?;
let readers = sorter.into_reader_cursors()?;
self.word_docids.extend(readers);
let sorter = exact_word_docids.into_sorter()?;
let readers = sorter.into_reader_cursors()?;
self.exact_word_docids.extend(readers);
let sorter = word_position_docids.into_sorter()?;
let readers = sorter.into_reader_cursors()?;
self.word_position_docids.extend(readers);
let sorter = fid_word_count_docids.into_sorter()?;
let readers = sorter.into_reader_cursors()?;
self.fid_word_count_docids.extend(readers);
let mut word_fid_docids_readers = Ok(vec![]);
let mut word_docids_readers = Ok(vec![]);
let mut exact_word_docids_readers = Ok(vec![]);
let mut word_position_docids_readers = Ok(vec![]);
let mut fid_word_count_docids_readers = Ok(vec![]);
rayon::scope(|s| {
s.spawn(|_| {
word_fid_docids_readers =
word_fid_docids.into_sorter().and_then(|s| s.into_reader_cursors());
});
s.spawn(|_| {
word_docids_readers =
word_docids.into_sorter().and_then(|s| s.into_reader_cursors());
});
s.spawn(|_| {
exact_word_docids_readers =
exact_word_docids.into_sorter().and_then(|s| s.into_reader_cursors());
});
s.spawn(|_| {
word_position_docids_readers =
word_position_docids.into_sorter().and_then(|s| s.into_reader_cursors());
});
s.spawn(|_| {
fid_word_count_docids_readers =
fid_word_count_docids.into_sorter().and_then(|s| s.into_reader_cursors());
});
});
self.word_fid_docids.extend(word_fid_docids_readers?);
self.word_docids.extend(word_docids_readers?);
self.exact_word_docids.extend(exact_word_docids_readers?);
self.word_position_docids.extend(word_position_docids_readers?);
self.fid_word_count_docids.extend(fid_word_count_docids_readers?);
Ok(())
}
@ -509,6 +526,10 @@ impl WordDocidsExtractors {
))
});
{
let span =
tracing::trace_span!(target: "indexing::documents::extract", "docids_extraction");
let _entered = span.enter();
document_changes.into_par_iter().try_for_each(|document_change| {
context_pool.with(|(rtxn, document_tokenizer, fields_ids_map, cached_sorter)| {
Self::extract_document_change(
@ -521,7 +542,12 @@ impl WordDocidsExtractors {
)
})
})?;
}
{
let span =
tracing::trace_span!(target: "indexing::documents::extract", "merger_building");
let _entered = span.enter();
let mut builder = WordDocidsMergerBuilders::new();
for (_rtxn, _tokenizer, _fields_ids_map, cache) in context_pool.into_items() {
builder.add_sorters(cache)?;
@ -529,6 +555,7 @@ impl WordDocidsExtractors {
Ok(builder.build())
}
}
fn extract_document_change(
rtxn: &RoTxn,

2
milli/src/update/new/extract/searchable/extract_word_pair_proximity_docids.rs
View File

@ -107,7 +107,7 @@ impl SearchableExtractor for WordPairProximityDocidsExtractor {
cached_sorter.insert_add_u32(key, docid)?;
}
}
};
}
}
Ok(())

29
milli/src/update/new/extract/searchable/mod.rs
View File

@ -13,7 +13,7 @@ pub use extract_word_docids::{
pub use extract_word_pair_proximity_docids::WordPairProximityDocidsExtractor;
use grenad::Merger;
use heed::RoTxn;
use rayon::iter::{IntoParallelIterator, ParallelIterator};
use rayon::iter::{IntoParallelIterator, ParallelBridge, ParallelIterator};
use tokenize_document::{tokenizer_builder, DocumentTokenizer};
use super::cache::CboCachedSorter;
@ -78,6 +78,10 @@ pub trait SearchableExtractor {
))
});
{
let span =
tracing::trace_span!(target: "indexing::documents::extract", "docids_extraction");
let _entered = span.enter();
document_changes.into_par_iter().try_for_each(|document_change| {
context_pool.with(|(rtxn, document_tokenizer, fields_ids_map, cached_sorter)| {
Self::extract_document_change(
@ -90,16 +94,27 @@ pub trait SearchableExtractor {
)
})
})?;
let mut builder = grenad::MergerBuilder::new(MergeDeladdCboRoaringBitmaps);
for (_rtxn, _tokenizer, _fields_ids_map, cache) in context_pool.into_items() {
let sorter = cache.into_sorter()?;
let readers = sorter.into_reader_cursors()?;
builder.extend(readers);
}
{
let mut builder = grenad::MergerBuilder::new(MergeDeladdCboRoaringBitmaps);
let span =
tracing::trace_span!(target: "indexing::documents::extract", "merger_building");
let _entered = span.enter();
let readers: Vec<_> = context_pool
.into_items()
.par_bridge()
.map(|(_rtxn, _tokenizer, _fields_ids_map, cached_sorter)| {
let sorter = cached_sorter.into_sorter()?;
sorter.into_reader_cursors()
})
.collect();
for reader in readers {
builder.extend(reader?);
}
Ok(builder.build())
}
}
fn extract_document_change(
rtxn: &RoTxn,