[
https://issues.apache.org/jira/browse/SOLR-9555?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15888501#comment-15888501
]
Mark Miller commented on SOLR-9555:
-----------------------------------
This is a great start Mike, I'll take a look. One thing I have been thinking
about is perhaps queuing up LIR publishes to ZK for a second or 2 rather than
hitting it for every update. You can index thousands of documents per second
and they can fail by the thousands, so it would be nice to have a little
throttle on ZK communication.
bq. Might need a flag to mark the node as dead locally or something like that.
I don't know that it's critical, because this would always be a problem to a
lesser degree anyway, but this is probably a good idea. I don't know how tricky
it ends up being, but seems like we could locally mark the state as down until
we notice it's state change in the clusterstate.
> Leader incorrectly publishes state for replica when it puts replica into LIR.
> -----------------------------------------------------------------------------
>
> Key: SOLR-9555
> URL: https://issues.apache.org/jira/browse/SOLR-9555
> Project: Solr
> Issue Type: Bug
> Security Level: Public(Default Security Level. Issues are Public)
> Reporter: Alan Woodward
> Attachments: SOLR-9555-WIP.patch
>
>
> See
> https://jenkins.thetaphi.de/job/Lucene-Solr-master-Linux/17888/consoleFull
> for an example
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
