We are experiencing issues with our jenkins server where it will regularly lock the build queue such that no jobs can be scheduled, and no jobs in the queue start running. At this point we are also unable to use the /script endpoint to clear the build queue

       

      We have been unable to determine the root cause of this issue, and restarting the jenkins server causes it to come up in a locked state as well. We have attempted to remove all the files that we know of that would tell Jenkins to resume jobs after a restart, but have been unable to find the comprehensive list. In particular are the pipeline jobs that come up and try to resume. Below is the list of files that we have tried removing: 

       

      /var/lib/jenkins/queue.xml
      /var/lib/jenkins/queue.xml.bak
      /var/lib/jenkins/org.jenkinsci.plugins.workflow.flow.FlowExecutionList.xml
      /var/lib/jenkins/org.jenkinsci.plugins.workflow.flow.FlowExecutionList.bak
      /var/lib/jenkins/org.jenkinsci.plugins.workflow.support.steps.StageStep.xml 
      

       

      Other triage steps we have taken: 

      • Increased JENKINS_HANDLER_MAX from 300 to 600 in /etc/sysconfig/jenkins
      • Starting in shutdown mode with a clear build queue
      • disabling jobs that queue frequently

      (first 3 edits were on the same day the ticket was created, Tuesday August 1, 2017)

      edit: 

      We bumped the number of executors to 2, and we see jobs queueing and running, but only on the second executor, not the first one. 

      edit 2: 

      Added an image from the /monitoring endpoint, the active thread count is the only sign that something has gone wrong. No other metric appears to be affected in a meaningful way

      edit 3:

      We managed to get a thread dump from the /threadDump endpoint. There 3 different suspect thread dumps: 

      "AtmostOneTaskExecutor[Periodic Jenkins queue maintenance] [#417]" Id=19597 Group=main WAITING on com.google.common.util.concurrent.AbstractFuture$Sync@488e43ab at sun.misc.Unsafe.park(Native Method) - waiting on com.google.common.util.concurrent.AbstractFuture$Sync@488e43ab at java.util.concurrent.locks.LockSupport.park(LockSupport.java:175) at java.util.concurrent.locks.AbstractQueuedSynchronizer.parkAndCheckInterrupt(AbstractQueuedSynchronizer.java:836) at java.util.concurrent.locks.AbstractQueuedSynchronizer.doAcquireSharedInterruptibly(AbstractQueuedSynchronizer.java:997) at java.util.concurrent.locks.AbstractQueuedSynchronizer.acquireSharedInterruptibly(AbstractQueuedSynchronizer.java:1304) at com.google.common.util.concurrent.AbstractFuture$Sync.get(AbstractFuture.java:275) at com.google.common.util.concurrent.AbstractFuture.get(AbstractFuture.java:111) at org.jenkinsci.plugins.workflow.cps.CpsStepContext.getThreadGroupSynchronously(CpsStepContext.java:248) at org.jenkinsci.plugins.workflow.cps.CpsStepContext.getThreadSynchronously(CpsStepContext.java:237) at org.jenkinsci.plugins.workflow.cps.CpsStepContext.doGet(CpsStepContext.java:294) at org.jenkinsci.plugins.workflow.support.DefaultStepContext.get(DefaultStepContext.java:61) at org.jenkinsci.plugins.workflow.support.steps.ExecutorStepExecution$PlaceholderTask.getNode(ExecutorStepExecution.java:259) at hudson.plugins.throttleconcurrents.ThrottleQueueTaskDispatcher.categoriesForPipeline(ThrottleQueueTaskDispatcher.java:411) at hudson.plugins.throttleconcurrents.ThrottleQueueTaskDispatcher.canRun(ThrottleQueueTaskDispatcher.java:168) at hudson.model.Queue.isBuildBlocked(Queue.java:1184) at hudson.model.Queue.maintain(Queue.java:1505) at hudson.model.Queue$1.call(Queue.java:320) at hudson.model.Queue$1.call(Queue.java:317) at jenkins.util.AtmostOneTaskExecutor$1.call(AtmostOneTaskExecutor.java:108) at jenkins.util.AtmostOneTaskExecutor$1.call(AtmostOneTaskExecutor.java:98) at jenkins.security.ImpersonatingExecutorService$2.call(ImpersonatingExecutorService.java:71) at java.util.concurrent.FutureTask.run(FutureTask.java:266) at hudson.remoting.AtmostOneThreadExecutor$Worker.run(AtmostOneThreadExecutor.java:110) at java.lang.Thread.run(Thread.java:748) Number of locked synchronizers = 1 - java.util.concurrent.locks.ReentrantLock$NonfairSync@529a7a50
      "Computer.threadPoolForRemoting [#80]" Id=19565 Group=main BLOCKED on hudson.model.Queue@770f2152 owned by "Computer.threadPoolForRemoting [#84]" Id=19573 at com.nirima.jenkins.plugins.docker.strategy.DockerOnceRetentionStrategy$1.run(DockerOnceRetentionStrategy.java:110) - blocked on hudson.model.Queue@770f2152 at jenkins.util.ContextResettingExecutorService$1.run(ContextResettingExecutorService.java:28) at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511) at java.util.concurrent.FutureTask.run(FutureTask.java:266) at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617) at java.lang.Thread.run(Thread.java:748) Number of locked synchronizers = 1 - java.util.concurrent.ThreadPoolExecutor$Worker@3677ff7e
      "Computer.threadPoolForRemoting [#81]" Id=19569 Group=main TIMED_WAITING on java.util.concurrent.SynchronousQueue$TransferStack@18c5c88 at sun.misc.Unsafe.park(Native Method) - waiting on java.util.concurrent.SynchronousQueue$TransferStack@18c5c88 at java.util.concurrent.locks.LockSupport.parkNanos(LockSupport.java:215) at java.util.concurrent.SynchronousQueue$TransferStack.awaitFulfill(SynchronousQueue.java:460) at java.util.concurrent.SynchronousQueue$TransferStack.transfer(SynchronousQueue.java:362) at java.util.concurrent.SynchronousQueue.poll(SynchronousQueue.java:941) at java.util.concurrent.ThreadPoolExecutor.getTask(ThreadPoolExecutor.java:1066) at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1127) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617) at java.lang.Thread.run(Thread.java:748)
      

       

      EDIT 4 (08-03-17|August 3): We seem to have stabilized the server.

      In our set up, a pipeline build that was interrupted would attempt to resume upon Jenkins restart. However when it does so it uses the exact node name that it had before which in this case is a docker container name with a unique hash that no longer exists. This causes the build to zombie itself and hang.

      The only indication on the file system that a pipeline job has not completed is that most recent stage xml file in the workflow directory for a given build does not contain a "<result>" tag. We wrote a python script that we ran on the jenkins master server to look for that and delete any build that did not have a "<result>" tag. We would appreciate any feed back on this approach, the python script is attached below. 

      The directory for a given job is: 

      /var/lib/jenkins/jobs/<JOB_NAME>/builds/<BUILD_NUMBER>/workflow

       

          [JENKINS-45917] [Jenkins v2.63] Build queue deadlocks

          integer Would switching to a different Retention strategy avoid this? I have to look at the Yet Another Docker Plugin to see what options exist that might emulate the behavior of provisioning and throwing away the container when done. Or do all the retention strategies have a lock on the build queue?  (I know this doesn't appear specific to the Docker plugins as tghe "Once Retention Strategy" is not just a Docker thing).

          Maxfield Stewart added a comment - integer Would switching to a different Retention strategy avoid this? I have to look at the Yet Another Docker Plugin to see what options exist that might emulate the behavior of provisioning and throwing away the container when done. Or do all the retention strategies have a lock on the build queue?  (I know this doesn't appear specific to the Docker plugins as tghe "Once Retention Strategy" is not just a Docker thing).

          oleg_nenashev We definitely use the TCB (Throttle Concurrent builds) plugin on both our Jenkisn environments (non Docker Plugin and Docker Plugin) which is a good lead into why we might see this behavior on our more traditional Jenkins setup as well. 

          If a pipeline is trying to resume AND has a throttle associated with it and is queued at the same time for another run, could that cause a deadlock? 

          Maxfield Stewart added a comment - oleg_nenashev We definitely use the TCB (Throttle Concurrent builds) plugin on both our Jenkisn environments (non Docker Plugin and Docker Plugin) which is a good lead into why we might see this behavior on our more traditional Jenkins setup as well.  If a pipeline is trying to resume AND has a throttle associated with it and is queued at the same time for another run, could that cause a deadlock? 

          As more info, last night before going home, we crawled all the job folders on our server looking for any job that had any historical message in it's history logs of "attempting to resume". If we found that log message, we purged that build history from disk.

           

          We restarted the jenkins server and have had 14 hours of consecutive up time (our previous high was about 2 hrs). Which is why we believe the culprit is the interaction with resume features of jenkins, but we're not sure exactly what causes Jenkins to believe a pipeline job needs to be resumed.  We know it's more than just purging the xml files in the jenkins home folder that control queue (queue.xml, org.jenkinsci.plugins.workflow.flow.FlowExecutionList.xml org.jenkinsci.plugins.workflow.support.steps.StageStep.xml ) because resumes still occur after clearing those.

          After doing our build history purge we appeared to get no resumes (and no locks)

          Maxfield Stewart added a comment - As more info, last night before going home, we crawled all the job folders on our server looking for any job that had any historical message in it's history logs of "attempting to resume". If we found that log message, we purged that build history from disk.   We restarted the jenkins server and have had 14 hours of consecutive up time (our previous high was about 2 hrs). Which is why we believe the culprit is the interaction with resume features of jenkins, but we're not sure exactly what causes Jenkins to believe a pipeline job needs to be resumed.  We know it's more than just purging the xml files in the jenkins home folder that control queue (queue.xml, org.jenkinsci.plugins.workflow.flow.FlowExecutionList.xml org.jenkinsci.plugins.workflow.support.steps.StageStep.xml ) because resumes still occur after clearing those. After doing our build history purge we appeared to get no resumes (and no locks)

          I also see ghprb plugin with some weird locks, maybe try disable it. 

          Kanstantsin Shautsou added a comment - I also see ghprb plugin with some weird locks, maybe try disable it. 

          We seem to have stabilized the server.

          In our set up, a pipeline build that was interrupted would attempt to resume upon Jenkins restart. However when it does so it uses the exact node name that it had before which in this case is a docker container name with a unique hash that no longer exists. This causes the build to zombie itself and hang.

          The only indication on the file system that a pipeline job has not completed is that most recent stage xml file in the workflow directory for a given build does not contain a "<result>" tag. We wrote a python script that we ran on the jenkins master server to look for that and delete any build that did not have a "<result>" tag. We would appreciate any feed back on this approach, the python script is attached to the ticket. 

          The directory for a given job is: 

          /var/lib/jenkins/jobs/<JOB_NAME>/builds/<BUILD_NUMBER>/workflow

          Marcos Pedreiro added a comment - We seem to have stabilized the server. In our set up, a pipeline build that was interrupted would attempt to resume upon Jenkins restart. However when it does so it uses the exact node name that it had before which in this case is a docker container name with a unique hash that no longer exists. This causes the build to zombie itself and hang. The only indication on the file system that a pipeline job has not completed is that most recent stage xml file in the workflow directory for a given build does not contain a "<result>" tag. We wrote a python script that we ran on the jenkins master server to look for that and delete any build that did not have a "<result>" tag. We would appreciate any feed back on this approach, the python script is attached to the ticket.  The directory for a given job is:  /var/lib/jenkins/jobs/<JOB_NAME>/builds/<BUILD_NUMBER>/workflow

          Oleg Nenashev added a comment -

          CC abayer and svanoort ^^^

          Oleg Nenashev added a comment - CC abayer and svanoort ^^^

          Sam Van Oort added a comment -

          Looks like a clone of https://issues.jenkins-ci.org/browse/JENKINS-36013 to me – I have a fix almost ready to go for that one

          Sam Van Oort added a comment - Looks like a clone of https://issues.jenkins-ci.org/browse/JENKINS-36013  to me – I have a fix almost ready to go for that one

          Daniel Beck added a comment -

          Does the fix for JENKINS-36013 address this?

          Daniel Beck added a comment - Does the fix for JENKINS-36013 address this?

          Sam Van Oort added a comment -

          danielbeck yes, it should - albeit after a 5 minute timeout before the build aborts when it can't resume on the same node (safety measure).  mpedreiro can you try installing the latest version of durable task step plugin and let us know if that doesn't settle it out (bear in mind that there is a delay before the build is killed when it cannot resume cleanly). 

          Thanks!

          Sam Van Oort added a comment - danielbeck  yes, it should - albeit after a 5 minute timeout before the build aborts when it can't resume on the same node (safety measure).  mpedreiro can you try installing the latest version of durable task step plugin and let us know if that doesn't settle it out (bear in mind that there is a delay before the build is killed when it cannot resume cleanly).  Thanks!

          Sam Van Oort added a comment -

          Closing as duplicate of the linked issue, since the fix there should cover this as well.

          Sam Van Oort added a comment - Closing as duplicate of the linked issue, since the fix there should cover this as well.

            Unassigned Unassigned
            mpedreiro Marcos Pedreiro
            Votes:
            0 Vote for this issue
            Watchers:
            5 Start watching this issue

              Created:
              Updated:
              Resolved: