[
https://issues.apache.org/jira/browse/HADOOP-6528?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12850346#action_12850346
]
Greg Wilkins commented on HADOOP-6528:
--------------------------------------
Hi,
we've been looking at the jetty code again and again and can't see how a
volatile or synchronized block or any memory barrier could help - unless it is
an issue within the JVM libraries themselves.
We've been running stop/start tests but also cannot reproduce.
It is possible that something else in hadoop is stopping the server before the
call to get the port?
Can we get full details of the platforms on which this fails - operating
system, JVM version etc. etc.
> Jetty returns -1 resulting in Hadoop masters / slaves to fail during startup.
> -----------------------------------------------------------------------------
>
> Key: HADOOP-6528
> URL: https://issues.apache.org/jira/browse/HADOOP-6528
> Project: Hadoop Common
> Issue Type: Bug
> Reporter: Hemanth Yamijala
> Attachments: jetty-server-failure.log
>
>
> A recent test failure on Hudson seems to indicate that Jetty's
> Server.getConnectors()[0].getLocalPort() is returning -1 in the
> HttpServer.getPort() method. When this happens, Hadoop masters / slaves that
> use Jetty fail to startup correctly.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
