You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Describe the bug
Significant performance drop in searches/aggregations when using Elasticsearch 7.7.0+ in Snow Owl 7.7.0+ release.
To Reproduce
Run a validation rule that uses either a search or aggregation with a large number of terms filter values (all active concept IDs listed, huge ECL/QL result set, etc.).
Expected behavior
Running any search with large number of term filter values should not cause significant performance drop in computation time.
Describe the bug
Significant performance drop in searches/aggregations when using Elasticsearch 7.7.0+ in Snow Owl 7.7.0+ release.
To Reproduce
Run a validation rule that uses either a search or aggregation with a large number of terms filter values (all active concept IDs listed, huge ECL/QL result set, etc.).
Expected behavior
Running any search with large number of term filter values should not cause significant performance drop in computation time.
Version
Additional context
Elasticsearch PR where the Percolator FetchPhase implementation rewrite happened:
elastic/elasticsearch#53728
Lucene 8.6.0 PR where the potential fix has been introduced:
apache/lucene-solr#1465
Related Elastic.co Discussion:
https://discuss.elastic.co/t/performance-problem-in-percolator-fetch-phase-classes/243748
This performance issue is going to be fixed when Elasticsearch 7.9.0 releases (it uses Lucene 8.6.0 under the hood).
The text was updated successfully, but these errors were encountered: