[
https://issues.apache.org/jira/browse/LUCENE-10165?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17427221#comment-17427221
]
Adrien Grand commented on LUCENE-10165:
---------------------------------------
The linked PR improved doc values merge times by 11% on the NYC taxis benchmark.
> Implement Lucene90DocValuesProducer#getMergeInstance
> ----------------------------------------------------
>
> Key: LUCENE-10165
> URL: https://issues.apache.org/jira/browse/LUCENE-10165
> Project: Lucene - Core
> Issue Type: Improvement
> Reporter: Adrien Grand
> Priority: Minor
> Time Spent: 10m
> Remaining Estimate: 0h
>
> The Lucene90 doc values producer optimizes for random access so that
> selective queries only have to decode the values that they need for sorting
> or faceting.
> However, in the case of merging, the merging process systematically consumes
> all doc IDs / values sequentially, so we could optimize for this access
> pattern via the merge instance?
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
