[
https://issues.apache.org/jira/browse/SOLR-7736?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16438035#comment-16438035
]
Tomás Fernández Löbbe commented on SOLR-7736:
---------------------------------------------
I think we should just return in this particular case. No need to continue
processing in that particular case (the {{updateLock.lockInterruptibly();}}
will generate another InterruptedException that will be logged (again) and
break the while, but no need to wait for that I think.
While looking at this particular class I noticed that other things my throw
InterruptedExceptions that we are just swallowing, like:
{code:java}
scheduledTriggers.add(entry.getValue());{code}
or
{code:java}
List<String> markers =
stateManager.listData(ZkStateReader.SOLR_AUTOSCALING_NODE_LOST_PATH);{code}
this throws {{Exception}} and we catch and log it, but that exception could
actually be an {{InterruptedException}}. Maybe we should change some of those
methods to throw exception types more specific than {{Exception}} (Including
{{InterruptedException}}), that way it will be harder to miss it
> Add a test for ZkController.publishAndWaitForDownStates
> -------------------------------------------------------
>
> Key: SOLR-7736
> URL: https://issues.apache.org/jira/browse/SOLR-7736
> Project: Solr
> Issue Type: Test
> Components: SolrCloud, Tests
> Reporter: Shalin Shekhar Mangar
> Assignee: Shalin Shekhar Mangar
> Priority: Minor
> Fix For: 7.4, master (8.0)
>
> Attachments: SOLR-7736.patch, SOLR-7736.patch,
> ZkController.failure.txt, consoleFull-2462-ZkControllerTest.txt.gz
>
>
> Add a test for ZkController.publishAndWaitForDownStates so that bugs like
> SOLR-6665 do not occur again. A test exists but it is not correct and
> currently disabled via AwaitsFix.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
