curator-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From John Vines <vi...@apache.org>
Subject Re: TestingServer randomly dying
Date Thu, 30 May 2013 20:54:00 GMT
Yeah, sorry for wasting your time on this one. I realized that it was stuck
in a different piece of code that was failing to hit zookeeper prior to
starting the TestingServer. My mistake.




On Thu, May 30, 2013 at 4:46 PM, Jordan Zimmerman <
jordan@jordanzimmerman.com> wrote:

> I'd need to see the tests to answer better, but make sure that the
> TestingServer is shutdown at the end of the test. Look at Curator's tests
> for good examples.
>
> -JZ
>
> On May 30, 2013, at 1:43 PM, John Vines <vines@apache.org> wrote:
>
> > I'm slowly adapting my codebase to use Curator, which includes using
> TestingServer's in a few places for junit tests. However, I've noticed an
> issue with my tests sometimes getting stuck in curator methods,
> specifically in the retry. In the output I see repeated Connection refused
> messages with occasional ConnectionLossException issues with a full stack
> trace. Netstatting shows the junit process no longer has a port bound, nor
> am I seeing any signs of life of zookeeper in a jstack. Unfortunately, I
> see nothing else in the output to indicate this thread dying.
> >
> > Is there any knowledge on this issue, either how to prevent it from
> happening or at least get a better diagnosis of what is going on? It's
> quite sporadic, which makes using this for a unit test quite painful.
> >
> > Thanks
> > John
>
>

Mime
View raw message