You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@lucene.apache.org by "David Smiley (Jira)" <ji...@apache.org> on 2021/03/05 17:55:00 UTC
[jira] [Commented] (SOLR-15215) SolrJ: Remove needless Netty
dependency
[ https://issues.apache.org/jira/browse/SOLR-15215?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17296241#comment-17296241 ]
David Smiley commented on SOLR-15215:
-------------------------------------
See [~janhoy] comment on the PR -- https://github.com/apache/lucene-solr/pull/2458#issuecomment-791505707
It was more worthy of greater visibility here in JIRA than code details there. Basically; Netty is needed if ZK is accessed via SSL. It was added in SOLR-13502
I think that we should not include it in SolrJ nonetheless. It may not be stated somewhere but I believe our strategic direction is for clients to not talk directly to ZooKeeper. We have Http2ClusterStateProvider for CloudSolrClient to use to talk to Solr to talk to ZK. It's not perfect (there are some perf bugs I've seen) but that's the direction. Not including Netty doesn't stop someone from adding it themselves. We could add some Javadoc comments in SolrJ CloudSolrClientBuilder on the ZK coordinates to mention the need for Netty when doing SSL. Not everyone uses SSL or perhaps some users do it differently like via [Istio service mesh|https://istio.io/latest/docs/concepts/security/].
> SolrJ: Remove needless Netty dependency
> ---------------------------------------
>
> Key: SOLR-15215
> URL: https://issues.apache.org/jira/browse/SOLR-15215
> Project: Solr
> Issue Type: Improvement
> Security Level: Public(Default Security Level. Issues are Public)
> Components: SolrJ
> Reporter: David Smiley
> Assignee: David Smiley
> Priority: Major
> Time Spent: 20m
> Remaining Estimate: 0h
>
> SolrJ depends on Netty transitively via ZooKeeper. But ZooKeeper's Netty dependency should be considered optional -- you have to opt-in.
> BTW it's only needed in Solr-core because of Hadoop/HDFS which ought to move to a contrib and take this dependency with it over there.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@lucene.apache.org
For additional commands, e-mail: issues-help@lucene.apache.org