Community
Participate
Working Groups
Dear Hudson manager, we have several jobs that are failing due to "All available display numbers are allocated" Extra tests build are failing for neon with the following message, did anyone ever got something similar? Checkout:workspace / /home/hudson/genie.papyrus/.hudson/jobs/Papyrus-Neon-Extra-Tests/workspace - hudson.remoting.LocalChannel@68f4bd45 Using strategy: Default Git Exe: git Last Built Revision: Revision 51c0baf5ff04ad6407acfd00cb245b5f48caf72d (origin/streams/2.0-maintenance) Checkout:source / /home/hudson/genie.papyrus/.hudson/jobs/Papyrus-Neon-Extra-Tests/workspace/source - hudson.remoting.LocalChannel@68f4bd45 Fetching changes from the remote Git repository Fetching upstream changes from /gitroot/papyrus/org.eclipse.papyrus.git Commencing build of Revision f9a89f1251398cfbdcf5d3bbc9f921ae25a2ea64 (origin/streams/2.0-maintenance) Checking out Revision f9a89f1251398cfbdcf5d3bbc9f921ae25a2ea64 (origin/streams/2.0-maintenance) Cleaning workspace FATAL: All available display numbers are allocated or blacklisted. allocated: [728, 729, 720, 721, 722, 723, 724, 725, 726, 727] blacklisted: [] java.lang.RuntimeException: All available display numbers are allocated or blacklisted. allocated: [728, 729, 720, 721, 722, 723, 724, 725, 726, 727] blacklisted: [] at hudson.plugins.xvnc.DisplayAllocator.doAllocate(DisplayAllocator.java:59) at hudson.plugins.xvnc.DisplayAllocator.allocate(DisplayAllocator.java:49) at hudson.plugins.xvnc.Xvnc.doSetUp(Xvnc.java:84) at hudson.plugins.xvnc.Xvnc.setUp(Xvnc.java:74) at hudson.model.Build$RunnerImpl.doRun(Build.java:129) at hudson.model.AbstractBuild$AbstractRunner.run(AbstractBuild.java:524) at hudson.model.Run.run(Run.java:1495) at hudson.model.FreeStyleBuild.run(FreeStyleBuild.java:44) at hudson.model.ResourceController.execute(ResourceController.java:82) at hudson.model.Executor.run(Executor.java:137)
I have reset the allocated display numbers. If this issue occurs more frequently there might be a bug in the xvnc plugin.
Hi, The problem seems to be back. It may be due to some admistrative task, cross-project is talking about "'hipp9' host was restarted" I restarted Papyrus HIPP but the problem is still here. See this failing job https://hudson.eclipse.org/papyrus/view/Sysml/job/papyrus-sysml-neon/191/console Can you reset allocated display numbers again? Regards,
Fixed.
(In reply to Frederic Gurr from comment #3) > Fixed. Yep its back, Thanks :)
Hi, The problem is once again back [1]. It has been some time since the last failure but I was wondering if, given your vision on the other projects and how often this might have happened, I should post it as a bug on the xvnc plugin ? Anyway could you reset the allocated numbers again please ? ^^' Thanks, Quentin [1] https://hudson.eclipse.org/papyrus/view/Oxygen%20(Master)/job/Papyrus-Master-Extra-Tests/lastBuild/console
Alternatively, if there is something we can do/run on our end each time we face this problem I would be happy to know it instead of asking for your intervention every time ;)
The job works again. Unfortunately there is nothing you can do on your end. This error should go away once the Papyrus HIPP is migrated to Jenkins (which uses a more recent version of the xvnc-plugin). In the meantime it needs to be fixed manually.
Yes it works again thanks :) Reading the internets I found out that what had to be done is to remove the locked numbers from the Xvnc.xml file but it had to be done after shuting down Hudson and needed an access to our Home. As I fear that the home/plugins are shared I was wondering if we had any access to that at all ^^'
Only admins have access to Hudson home directories. Sorry.
Hi, the problem seems to have re-manifest itself https://hudson.eclipse.org/tracecompass/job/tracecompass-gerrit/14408/console https://hudson.eclipse.org/tracecompass/job/tracecompass-gerrit/14407/console https://hudson.eclipse.org/tracecompass/job/tracecompass-gerrit/14406/console https://hudson.eclipse.org/tracecompass/job/tracecompass-gerrit/14405/console It seems to be only on hipp3 which was reset last night, maybe some lock file is lingering or something?
Different HIPP, but similar problem. It's fixed now.
I again, Same problem that fails multiple builds. We'll try to move to jenkins this year as soon as possible but in the meantime could you please fix it ? Thanks, Quentin
Fixed again.