lucene-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Mark Miller (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (SOLR-9826) Shutting down leader when it's sending updates makes another active node go into recovery
Date Wed, 07 Dec 2016 06:19:58 GMT

    [ https://issues.apache.org/jira/browse/SOLR-9826?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15727854#comment-15727854
] 

Mark Miller commented on SOLR-9826:
-----------------------------------

We should probably either stop interrupting the update executor on shutdown or special case
an interruption so that the leader doesn't consider it a failed update to a replica. We probably
want to fail the update to the user in that case. 

> Shutting down leader when it's sending updates makes another active node go into recovery
> -----------------------------------------------------------------------------------------
>
>                 Key: SOLR-9826
>                 URL: https://issues.apache.org/jira/browse/SOLR-9826
>             Project: Solr
>          Issue Type: Bug
>      Security Level: Public(Default Security Level. Issues are Public) 
>    Affects Versions: 6.3
>            Reporter: Ere Maijala
>              Labels: solrcloud
>         Attachments: failure.log
>
>
> If the leader in SolrCloud is sending updates to a follower when it's shut down, it forces
the replica it can't communicate with (due to being shut down, I assume) to go into recovery.
I'll attach a log excerpt that shows the related messages.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lucene.apache.org
For additional commands, e-mail: dev-help@lucene.apache.org


Mime
View raw message