[
https://issues.apache.org/jira/browse/SOLR-16061?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17505295#comment-17505295
]
David Smiley commented on SOLR-16061:
-------------------------------------
I believe I addressed all remaining issues.
If there is a mistake/error, it would be in some tests that mistakenly assume
that all solrClients are equivalent, when in some very small percentage of
scenarios, there is some timing issue of which ZkStateReader in which client is
being consulted. When the original work was done, had an IntelliJ "structural
replace" been done, I don't think such mistakes would have been made. It's too
tedious to review the massive PR on a file by file basis, although I've done
some of it. Some commit-by-commit that don't touch too many are useful; for
example I just added a convenience method on
MiniSolrCloudCluster.getZkStateReader which is wanted very often. At this
point, the tests pass!
> Decouple CloudSolrClient from ZkStateReader
> -------------------------------------------
>
> Key: SOLR-16061
> URL: https://issues.apache.org/jira/browse/SOLR-16061
> Project: Solr
> Issue Type: Task
> Security Level: Public(Default Security Level. Issues are Public)
> Reporter: Jan Høydahl
> Priority: Blocker
> Fix For: 9.0
>
> Time Spent: 2h 50m
> Remaining Estimate: 0h
>
> Spinoff from SOLR-15342
> We need to decouple CloudSolrClient from Zookeeper, to prepare for a future
> modularization of solrj. This should be targetet for 9.0.
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
