[
https://issues.apache.org/jira/browse/SOLR-13132?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17071103#comment-17071103
]
Michael Gibney commented on SOLR-13132:
---------------------------------------
Quick prose follow-up to March 16 patch version, pushed to [PR
#751|https://github.com/apache/lucene-solr/pull/751]. As I re-read the comment
above, I realize that a lot of the feeback applies to the caching change (now
excised and introduced separately under SOLR-13807).
The changes presented here now consist of:
# Refactoring facet collection in {{FacetFieldProcessorByArrayDV}} and
{{FacetFieldProcessorByArrayUIF/UnInvertedField}} to perform facet count
collection via the new {{SweepDISI}} abstraction. This refactoring is exercised
by every instance of facet count collection over these implementations, and
should indeed be well-covered by existing tests.
# A change in the implementation of {{RelatednessAgg}} which, for
{{FacetFieldProcessorByArray}} bulk collection over fields of cardinality > 1
(i.e., not refinement requests), leverages the new {{SweepDISI}} abstraction to
collect over multiple DocSet domains in a single sweep. No caching or
heuristics are involved, so this collection always takes place in exactly the
same way for a given input.
> Improve JSON "terms" facet performance when sorted by relatedness
> ------------------------------------------------------------------
>
> Key: SOLR-13132
> URL: https://issues.apache.org/jira/browse/SOLR-13132
> Project: Solr
> Issue Type: Improvement
> Components: Facet Module
> Affects Versions: 7.4, master (9.0)
> Reporter: Michael Gibney
> Priority: Major
> Attachments: SOLR-13132-with-cache-01.patch,
> SOLR-13132-with-cache.patch, SOLR-13132.patch
>
> Time Spent: 1.5h
> Remaining Estimate: 0h
>
> When sorting buckets by {{relatedness}}, JSON "terms" facet must calculate
> {{relatedness}} for every term.
> The current implementation uses a standard uninverted approach (either
> {{docValues}} or {{UnInvertedField}}) to get facet counts over the domain
> base docSet, and then uses that initial pass as a pre-filter for a
> second-pass, inverted approach of fetching docSets for each relevant term
> (i.e., {{count > minCount}}?) and calculating intersection size of those sets
> with the domain base docSet.
> Over high-cardinality fields, the overhead of per-term docSet creation and
> set intersection operations increases request latency to the point where
> relatedness sort may not be usable in practice (for my use case, even after
> applying the patch for SOLR-13108, for a field with ~220k unique terms per
> core, QTime for high-cardinality domain docSets were, e.g.: cardinality
> 1816684=9000ms, cardinality 5032902=18000ms).
> The attached patch brings the above example QTimes down to a manageable
> ~300ms and ~250ms respectively. The approach calculates uninverted facet
> counts over domain base, foreground, and background docSets in parallel in a
> single pass. This allows us to take advantage of the efficiencies built into
> the standard uninverted {{FacetFieldProcessorByArray[DV|UIF]}}), and avoids
> the per-term docSet creation and set intersection overhead.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@lucene.apache.org
For additional commands, e-mail: issues-h...@lucene.apache.org
