[
https://issues.apache.org/jira/browse/SOLR-9310?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15381495#comment-15381495
]
Pushkar Raste commented on SOLR-9310:
-------------------------------------
I was able to figure a few more things about what going own. The existing
{{PeerSyncTest}} does not change core's state from {{ACTIVE}} to recovering and
hence the condition following condition (block) in
{{DistributedUpdateLogProcessor.versionAdd()}} does not get executed
{code}
if (ulog.getState() != UpdateLog.State.ACTIVE && (cmd.getFlags() &
UpdateCommand.REPLAY) == 0) {
// we're not in an active state, and this update isn't from a
replay, so buffer it.
cmd.setFlags(cmd.getFlags() | UpdateCommand.BUFFERING);
ulog.add(cmd);
return true;
}
{code}
However, the test I have attached certainly goes through above mentioned code
which results in updates came from log replay to be buffered. As a result of
this {{compareFingerPrint()}} check at the end of {{PeerSync.handleUpdates()}}
fails,whenever a PeerSync was triggered and core was in not ACTIVE state. I am
not entirely sure why a core would in ACTIVE state if PeerSync was triggered
(which happens in {{PeerSyncTest}}).
I think {{compareFingerPrint}} check should be moved out of {{PeerSync}} class
to {{RecoveryStrategy}} after buffered updates are applied. It might also be a
good idea to move the commit after log replay as current commit seems to be
resulting in a NOOP.
> PeerSync fails on a node restart due to IndexFingerPrint mismatch
> -----------------------------------------------------------------
>
> Key: SOLR-9310
> URL: https://issues.apache.org/jira/browse/SOLR-9310
> Project: Solr
> Issue Type: Bug
> Security Level: Public(Default Security Level. Issues are Public)
> Reporter: Pushkar Raste
> Attachments: PeerSyncReplicationTest.patch
>
>
> I found that Peer Sync fails if a node restarts and documents were indexed
> while node was down. IndexFingerPrint check fails after recovering node
> applies updates.
> This happens only when node restarts and not if node just misses updates due
> reason other than it being down.
> Please check attached patch for the test.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org
