[
https://issues.apache.org/jira/browse/ZOOKEEPER-1852?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13858861#comment-13858861
]
Germán Blanco commented on ZOOKEEPER-1852:
------------------------------------------
So this is covering a problem when TruncateTest fails and that leaves a running
server that affects additional test cases later on, right?
I see now the goal, thank you for explaining.
+1
Would you mind taking a look at the patch (wrongly) in ZOOKEEPER-1833?
I didn't put it in a subtask by mistake, and I don't know if it makes sense to
do it now.
Anyway, I believe it solves a problem of a run-away server like this one, but
one that runs away in all cases (failure or success).
> ServerCnxnFactory instance is not properly cleanedup
> ----------------------------------------------------
>
> Key: ZOOKEEPER-1852
> URL: https://issues.apache.org/jira/browse/ZOOKEEPER-1852
> Project: ZooKeeper
> Issue Type: Sub-task
> Components: tests
> Reporter: Rakesh R
> Assignee: Rakesh R
> Fix For: 3.4.6, 3.5.0
>
> Attachments: ZOOKEEPER-1852-br-3.4.patch,
> ZOOKEEPER-1852-br-3.4.patch, ZOOKEEPER-1852.patch, ZOOKEEPER-1852.patch
>
>
> ClientBase#createNewServerInstance() - Say the startup of the server fails,
> this will not initialize 'serverFactory' and will be null. When the flow
> comes to teardown/shutdown, it will bypass stopping of this server instance
> due to the following check. This will affect other test case verifications
> like, jmx check 'JMXEnv#ensureOnly'.
> ClientBase#shutdownServerInstance
> {code}
> static void shutdownServerInstance(ServerCnxnFactory factory,
> String hostPort)
> {
> if (factory != null) {
> //...shutdown logic
> }
> {code}
--
This message was sent by Atlassian JIRA
(v6.1.5#6160)
