Details
-
Improvement
-
Status: Resolved
-
Major
-
Resolution: Fixed
-
None
-
None
-
None
-
New
Description
For the "browse facets" use case, where logically you run MatchAllDocsQuery and then compute facet hits, we can optimize this case for both SortedSetDocValuesFacets and FastTaxonomyFacetCounts so that we don't use the query DISI at all and rather just pull from the doc values iterator using nextDoc.