[
https://issues.apache.org/jira/browse/SOLR-9608?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15551305#comment-15551305
]
Mikhail Khludnev commented on SOLR-9608:
----------------------------------------
I'm fighting with it in branch_6x, in scope of SOLR-9604. The problem is
{{LBHttpSolrClient.makeSolrClient(String)}} uses {{HttpSolrClients}} as a
napkins. It creates it, use, and then throw client away. It's fine, until we
start to use HttpSolrClient identity as a user token to allow connection reuse.
Thus, we either need to manage LBHttpSolrClient to reuse HttpSolrClient
instances or come up with another user tokens. Was there any drawback in
initial [~romseygeek]'s idea to use a constant for that?
> ConnectionReuseTest failures
> ----------------------------
>
> Key: SOLR-9608
> URL: https://issues.apache.org/jira/browse/SOLR-9608
> Project: Solr
> Issue Type: Bug
> Security Level: Public(Default Security Level. Issues are Public)
> Reporter: Steve Rowe
> Assignee: Alan Woodward
>
> My Jenkins has seen this test fail three times in the last couple hours, all
> three reproduce for me - this is new, last failures were from May - here's
> the first failure:
> {noformat}
> Checking out Revision 0eb6b1c823d347319cc0894b5fea95f085d4c8d4
> (refs/remotes/origin/master)
> [...]
> [junit4] 2> NOTE: reproduce with: ant test
> -Dtestcase=ConnectionReuseTest -Dtests.method=testConnectionReuse
> -Dtests.seed=FA03A6C483303E7D -Dtests.slow=true -Dtests.locale=ga
> -Dtests.timezone=Etc/GMT-9 -Dtests.asserts=true -Dtests.file.encoding=UTF-8
> [junit4] FAILURE 1.09s J9 | ConnectionReuseTest.testConnectionReuse <<<
> [junit4] > Throwable #1: java.lang.AssertionError: We expected all
> communication to use one connection! HttpSolrClient 1
> [junit4] > at
> __randomizedtesting.SeedInfo.seed([FA03A6C483303E7D:DB3BCDAAAB786730]:0)
> [junit4] > at
> org.apache.solr.client.solrj.ConnectionReuseTest.testConnectionReuse(ConnectionReuseTest.java:158)
> {noformat}
> This test failed for me twice without specifying a seed, so it's likely
> failing (close to) 100% of the time.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org
