[
https://issues.apache.org/jira/browse/SOLR-14098?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16997464#comment-16997464
]
Yonik Seeley commented on SOLR-14098:
-------------------------------------
Tracing it back, it looks like this bug has been around since the
implementation of shard splitting: hash 873b2308b6f (SOLR-3755)
I don't have any idea if it actually manifests as bad behavior (probably not?),
but we should make the behavior consistent anyway.
> inconsistent replica state treatment in RequestApplyUpdatesOp
> -------------------------------------------------------------
>
> Key: SOLR-14098
> URL: https://issues.apache.org/jira/browse/SOLR-14098
> Project: Solr
> Issue Type: Bug
> Security Level: Public(Default Security Level. Issues are Public)
> Reporter: Yonik Seeley
> Priority: Major
>
> IN RequestApplyUpdatesOp, if there are no buffered updates, the method
> returns. If there are buffered updates, they are applied and the replica
> state is set to ACTIVE.
> I'm not sure what behavior is correct, but the ending replica state should
> obviously not depend on if there were buffered updates or not.
--
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
