[
https://issues.apache.org/jira/browse/SOLR-15412?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17347634#comment-17347634
]
Jan Høydahl edited comment on SOLR-15412 at 5/19/21, 12:44 PM:
---------------------------------------------------------------
But that's "A Good Thing™", isn't it? In a 10-node cluster, you start upgrading
one node, it will not start, so you abort there and investigate...? It would be
far worse if you completed the whole upgrade process and then suddently things
came crashing down?
NOTE I'm not advocating for not improving the validation logic, just trying to
understand how this is seen in practice during an upgrade.
was (Author: janhoy):
But that's "A Good Thing™", isn't it? In a 10-node cluster, you start upgrading
one node, it will not start, so you abort there and investigate...? It would be
far worse if you completed the whole upgrade process and then suddently things
came crashing down?
> Strict validation on Replica metadata can cause complete outage
> ---------------------------------------------------------------
>
> Key: SOLR-15412
> URL: https://issues.apache.org/jira/browse/SOLR-15412
> Project: Solr
> Issue Type: Bug
> Security Level: Public(Default Security Level. Issues are Public)
> Affects Versions: 8.5
> Reporter: Ishan Chattopadhyaya
> Priority: Blocker
> Fix For: 8.9
>
>
> More details in SOLR-14245. Intention of this issue is to reduce the impact
> of the changes in SOLR-14245 and immediately issue a breakfix release for the
> users who want to upgrade to 8.5+.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
