[jira] [Commented] (GEODE-7237) CI failure: ConnectCommandAcceptanceTest.invalidHostname

["Kirk Lund (Jira)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/GEODE-7237?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16937165#comment-16937165 ] 

Kirk Lund commented on GEODE-7237:
----------------------------------

This is caused by a race condition in the implementation of GfshRule. After forking the process, ProcessLogger is created to read the output from that process. ProcessLogger uses log4j-core to reconfigure logging. Meanwhile, the process may have already started printing output which has been missed which causes the assertion in this bug to fail.

I cannot find any good reason for ProcessLogger to actually need to use log4j-core to reconfigure logging. If I change it to just grab a logger without reconfiguring log4j, the test runs must faster and passes consistently.

> CI failure: ConnectCommandAcceptanceTest.invalidHostname
> --------------------------------------------------------
>
>                 Key: GEODE-7237
>                 URL: https://issues.apache.org/jira/browse/GEODE-7237
>             Project: Geode
>          Issue Type: Bug
>          Components: ci, tests
>            Reporter: Aaron Lindsey
>            Assignee: Kirk Lund
>            Priority: Major
>
> [https://concourse.apachegeode-ci.info/teams/main/pipelines/apache-develop-main/jobs/AcceptanceTestOpenJDK8/builds/1101]
> {code:java}
> org.apache.geode.management.internal.cli.commands.ConnectCommandAcceptanceTest > invalidHostname FAILED
>     java.lang.AssertionError: 
>     Expecting:
>      <"">
>     to contain:
>      <"can't be reached. Hostname or IP address could not be found."> 
>         at org.apache.geode.management.internal.cli.commands.ConnectCommandAcceptanceTest.invalidHostname(ConnectCommandAcceptanceTest.java:59)
> {code}



--
This message was sent by Atlassian Jira
(v8.3.4#803005)