river-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Greg Trasuk <tras...@stratuscom.com>
Subject Re: Tests failing due to port in use
Date Wed, 10 Apr 2013 11:39:44 GMT

I'm testing on JDK7 on a mac right now.  When I see a good test run
(started one last night.  Looks promising at the moment) I was going to
fire up tests on JDK6/Solaris 10, JDK7/win7x64 and JDK7/winxp32.

Given that the Levels fix is for a JDK7 issue, it's pretty critical that
we test on JDK7.  Even having said that, I don't know that the Levels
issue is immediately apparent - we should probably write a test that
demonstrates the issue and fix.

Is Dennis there?  Do you have some code that shows the Levels problem? 
If so I'll see about writing a jtreg test for it.



On Wed, 2013-04-10 at 07:12, Peter Firmstone wrote:
> Greg,
> Are you testing on Java 7?
> The last release was only tested with Java 5 and 6.
> Can you test with Java 6?
> Regards,
> Peter.
> On 10/04/2013 6:32 AM, Greg Trasuk wrote:
> > On Tue, 2013-04-09 at 15:29, Peter Firmstone wrote:
> >> This was being discussed in two separate threads:
> >>
> >> Greg,
> >>
> >> Are you able to find which process is using a port?  Perhaps jvisualvm
> >> might help determine which test it is?  If you can narrow it down to a
> >> bunch of test definitions sharing a test class, by identifying the class
> >> on the stack, you'll find it.
> >>
> > No, there's no process using the port.  The evidence points to a
> > configuration error.  The problem is not that the registrars don't
> > start.  It's that they start on random ports when the port is supposed
> > to be specified.

View raw message