searchengines-web/MeiliSearch
searchengines-web/MeiliSearch
1
0
Fork 0
mirror of https://github.com/meilisearch/MeiliSearch synced 2024-11-30 08:44:27 +01:00
Code Issues Releases Wiki Activity

Reduce the ram consumption when re-indexing all the documents

Browse Source
This commit is contained in:
Clément Renault 2019-10-29 17:46:23 +01:00
parent 41065305aa
commit 4dff8a215e
No known key found for this signature in database
GPG Key ID: 92ADA4E935E71FA4
1 changed files with 43 additions and 39 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

24
meilidb-core/src/update/documents_addition.rs
View File

@ -150,7 +150,7 @@ pub fn apply_documents_addition(
main_store,
postings_lists_store,
docs_words_store,
ranked_map,
&ranked_map,
number_of_inserted_documents,
indexer,
)
@ -179,24 +179,25 @@ pub fn reindex_all_documents(
}
// 2. remove the documents posting lists
let number_of_inserted_documents = documents_ids_to_reindex.len();
main_store.put_words_fst(writer, &fst::Set::default())?;
main_store.put_ranked_map(writer, &ranked_map)?;
main_store.put_number_of_documents(writer, |_| 0)?;
postings_lists_store.clear(writer)?;
docs_words_store.clear(writer)?;
// 3. re-index chunks of documents (otherwise we make the borrow checker unhappy)
for documents_ids in documents_ids_to_reindex.chunks(100) {
let stop_words = match main_store.stop_words_fst(writer)? {
Some(stop_words) => stop_words,
None => fst::Set::default(),
};
// 3. re-index one document by one document (otherwise we make the borrow checker unhappy)
let number_of_inserted_documents = documents_ids.len();
let mut indexer = RawIndexer::new(stop_words);
let mut ram_store = HashMap::new();
for document_id in documents_ids_to_reindex {
for result in documents_fields_store.document_fields(writer, document_id)? {
for document_id in documents_ids {
for result in documents_fields_store.document_fields(writer, *document_id)? {
let (attr, bytes) = result?;
let value: serde_json::Value = serde_json::from_slice(bytes)?;
ram_store.insert((document_id, attr), value);
@ -207,7 +208,7 @@ pub fn reindex_all_documents(
writer,
attr,
schema.props(attr),
docid,
*docid,
documents_fields_store,
documents_fields_counts_store,
&mut indexer,
@ -223,10 +224,13 @@ pub fn reindex_all_documents(
main_store,
postings_lists_store,
docs_words_store,
ranked_map,
&ranked_map,
number_of_inserted_documents,
indexer,
)
)?;
}
Ok(())
}
pub fn write_documents_addition_index(
@ -234,7 +238,7 @@ pub fn write_documents_addition_index(
main_store: store::Main,
postings_lists_store: store::PostingsLists,
docs_words_store: store::DocsWords,
ranked_map: RankedMap,
ranked_map: &RankedMap,
number_of_inserted_documents: usize,
indexer: RawIndexer,
) -> MResult<()> {
@ -279,7 +283,7 @@ pub fn write_documents_addition_index(
};
main_store.put_words_fst(writer, &words)?;
main_store.put_ranked_map(writer, &ranked_map)?;
main_store.put_ranked_map(writer, ranked_map)?;
main_store.put_number_of_documents(writer, |old| old + number_of_inserted_documents as u64)?;
Ok(())