We just upgraded to 2.129. Our jenkins is configured to start nodes as needed; release nodes as upon idle; and run a job on master which turns nodes off if they aren't in use.
We had the accelerated-build-now-plugin (version 1.0.1) plugin enabled.
When our normal or pipeline jobs start, they try to pick any node other than master. Jenkins starts the node, it appears in the list of available nodes, but Jenkins doesn't manage to start a build on the node:
SEVERE: Timer task hudson.model.Queue$MaintainTask@2041414e failed
java.lang.ClassCastException: org.jenkinsci.plugins.workflow.support.steps.ExecutorStepExecution$PlaceholderTask cannot be cast to hudson.model.AbstractProject
at org.terracotta.jenkins.plugins.acceleratedbuildnow.AcceleratedBuildNowComparator.compare(AcceleratedBuildNowComparator.java:21)
at org.terracotta.jenkins.plugins.acceleratedbuildnow.AcceleratedBuildNowComparator.compare(AcceleratedBuildNowComparator.java:11)
at java.util.TimSort.countRunAndMakeAscending(TimSort.java:355)
at java.util.TimSort.sort(TimSort.java:220)
at java.util.Arrays.sort(Arrays.java:1512)
at java.util.ArrayList.sort(ArrayList.java:1462)
at java.util.Collections.sort(Collections.java:175)
at org.terracotta.jenkins.plugins.acceleratedbuildnow.AcceleratedBuildNowSorter.sortBuildableItems(AcceleratedBuildNowSorter.java:30)
at hudson.model.Queue.maintain(Queue.java:1571)
at hudson.model.Queue$MaintainTask.doRun(Queue.java:2862)
at hudson.triggers.SafeTimerTask.run(SafeTimerTask.java:72)
at jenkins.security.ImpersonatingScheduledExecutorService$1.run(ImpersonatingScheduledExecutorService.java:58)
at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
at java.util.concurrent.FutureTask.runAndReset(FutureTask.java:308)
at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$301(ScheduledThreadPoolExecutor.java:180)
at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:294)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
at java.lang.Thread.run(Thread.java:748)
Instead, the jobs sit waiting, and eventually the computers being idle are released by jenkins and then reaped by the job on master.
Disabling this plugin results in our system working as it used to work.
Originally reported by jsoref, imported from: Jenkins Queue DoS when QueueItemSorter extension throws runtime Exception
- assignee:
anthonydahanne
- status: Resolved
- priority: Blocker
- component(s): core
- label(s): 2.121.3-fixed
- resolution: Fixed
- resolved: 2018-07-03T18:01:45+00:00
- votes: 0
- watchers: 3
- imported: 2025-11-24
Raw content of original issue
We just upgraded to 2.129. Our jenkins is configured to start nodes as needed; release nodes as upon idle; and run a job on master which turns nodes off if they aren't in use.
We had the accelerated-build-now-plugin (version 1.0.1) plugin enabled.
When our normal or pipeline jobs start, they try to pick any node other than master. Jenkins starts the node, it appears in the list of available nodes, but Jenkins doesn't manage to start a build on the node:
SEVERE: Timer task hudson.model.Queue$MaintainTask@2041414e failed
java.lang.ClassCastException: org.jenkinsci.plugins.workflow.support.steps.ExecutorStepExecution$PlaceholderTask cannot be cast to hudson.model.AbstractProject
at org.terracotta.jenkins.plugins.acceleratedbuildnow.AcceleratedBuildNowComparator.compare(AcceleratedBuildNowComparator.java:21)
at org.terracotta.jenkins.plugins.acceleratedbuildnow.AcceleratedBuildNowComparator.compare(AcceleratedBuildNowComparator.java:11)
at java.util.TimSort.countRunAndMakeAscending(TimSort.java:355)
at java.util.TimSort.sort(TimSort.java:220)
at java.util.Arrays.sort(Arrays.java:1512)
at java.util.ArrayList.sort(ArrayList.java:1462)
at java.util.Collections.sort(Collections.java:175)
at org.terracotta.jenkins.plugins.acceleratedbuildnow.AcceleratedBuildNowSorter.sortBuildableItems(AcceleratedBuildNowSorter.java:30)
at hudson.model.Queue.maintain(Queue.java:1571)
at hudson.model.Queue$MaintainTask.doRun(Queue.java:2862)
at hudson.triggers.SafeTimerTask.run(SafeTimerTask.java:72)
at jenkins.security.ImpersonatingScheduledExecutorService$1.run(ImpersonatingScheduledExecutorService.java:58)
at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
at java.util.concurrent.FutureTask.runAndReset(FutureTask.java:308)
at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$301(ScheduledThreadPoolExecutor.java:180)
at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:294)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
at java.lang.Thread.run(Thread.java:748)
Instead, the jobs sit waiting, and eventually the computers being idle are released by jenkins and then reaped by the job on master.
Disabling this plugin results in our system working as it used to work.
We just upgraded to 2.129. Our jenkins is configured to start nodes as needed; release nodes as upon idle; and run a job on master which turns nodes off if they aren't in use.
We had the accelerated-build-now-plugin (version 1.0.1) plugin enabled.
When our normal or pipeline jobs start, they try to pick any node other than master. Jenkins starts the node, it appears in the list of available nodes, but Jenkins doesn't manage to start a build on the node:
Instead, the jobs sit waiting, and eventually the computers being idle are released by jenkins and then reaped by the job on master.
Disabling this plugin results in our system working as it used to work.
Originally reported by jsoref, imported from: Jenkins Queue DoS when QueueItemSorter extension throws runtime Exception
Raw content of original issue