[
https://issues.apache.org/jira/browse/SOLR-14342?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
David Smiley updated SOLR-14342:
--------------------------------
Labels: scaling (was: )
> CoreSorter is partially broken, thus core loading order is sub-optimal
> ----------------------------------------------------------------------
>
> Key: SOLR-14342
> URL: https://issues.apache.org/jira/browse/SOLR-14342
> Project: Solr
> Issue Type: Improvement
> Security Level: Public(Default Security Level. Issues are Public)
> Components: SolrCloud
> Reporter: David Smiley
> Assignee: David Smiley
> Priority: Major
> Labels: scaling
> Time Spent: 1h 50m
> Remaining Estimate: 0h
>
> In SOLR-7280, it Solr was supposedly improved to load cores in a more optimal
> order in SolrCloud, considering the state of collections across the cluster.
> The CoreContainer uses CoreSorter for this. Unfortunately, CoreSorter's
> attempts to gather statistics early on in CoreSorter.init() operate on an
> empty list and do nothing (i.e. dead code). The root cause is that
> getCloudDescriptors() works by looking at cc.getCores however at this stage
> _there are no cores_ ! (We haven't sorted them so they certainly haven't been
> loaded yet. I have a fix.
> This might be classified as a bug fix but the core load order is more about
> optimization, so a fix to a wrong/incomplete optimization isn't really a bug
> from a user's point of view.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@lucene.apache.org
For additional commands, e-mail: issues-h...@lucene.apache.org
