[
https://issues.apache.org/jira/browse/SOLR-8367?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15047731#comment-15047731
]
Mike Drob commented on SOLR-8367:
---------------------------------
SyncStrategy, the try-catch at line ~218 can go away.
Do we need to check {{isClosed}} again during {{requestRecoveries}}? It's
possible that it's false when the recoveries are being set up, but changes to
true later by the time we actually make the RPC.
An optimization might be to break out of the whole loop when closed, since it
looks like not much will be happening anyway.
In ZkController, can we log which retry we are on (in both places)? That will
make parsing logs easier when this failure happens.
You have a couple {{System.out.println}} in SolrCore that could probably be
{{log.debug}} or even {{trace}}.
In the test, you could store {{(HttpSolrClient) clients.get(0)).getBaseURL()}}
as a local variable instead of loading it twice.
If you're fixing shard count to 2 in the test, do we still want to
{{createCollection(testCollectionName, 1, 3, 1);}} for three shards?
Architecture question -- What happens when you send an update FROMLEADER to the
one that happens to be the leader? Also, why are we using decreasing versions?
> The new LIR 'all replicas participate' failsafe code needs to be improved.
> --------------------------------------------------------------------------
>
> Key: SOLR-8367
> URL: https://issues.apache.org/jira/browse/SOLR-8367
> Project: Solr
> Issue Type: Bug
> Reporter: Mark Miller
> Assignee: Mark Miller
> Attachments: SOLR-8367.patch, SOLR-8367.patch
>
>
> For one, it currently only kicks in the first attempted leader. If it's
> another replica that is stuck in LIR, it won't help.
> Second, when we attempt to be leader, knowing we might fail due to LIR, we
> should not put other replicas into recovery if they fail to sync with us -
> not until we know we will actually be leader.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
