[
https://issues.apache.org/jira/browse/LUCENE-5644?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Michael McCandless reopened LUCENE-5644:
----------------------------------------
Reopening ...
I realized the simple LIFO logic is too simple, right after a flush. When that
happens, we should try to pick a ThreadState that's already initialized, so
that if no full-flushing (getReader) is happening, we don't tie up RAM
indefinitely in the pending segments.
> ThreadAffinityDocumentsWriterThreadPool should clear the bindings on flush
> --------------------------------------------------------------------------
>
> Key: LUCENE-5644
> URL: https://issues.apache.org/jira/browse/LUCENE-5644
> Project: Lucene - Core
> Issue Type: Bug
> Components: core/index
> Reporter: Michael McCandless
> Assignee: Michael McCandless
> Fix For: 4.8.1, 4.9, 5.0
>
> Attachments: LUCENE-5644.patch, LUCENE-5644.patch, LUCENE-5644.patch,
> LUCENE-5644.patch, LUCENE-5644.patch
>
>
> This class remembers which thread used which DWPT, but it never clears
> this "affinity". It really should clear it on flush, this way if the
> number of threads doing indexing has changed we only use as many DWPTs
> as there are incoming threads.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org
