Details
-
Bug
-
Status: Closed (View Workflow)
-
Major
-
Resolution: Fixed
-
None
-
-
indian
Description
There is a multibranch job taht bwalding runs that fails to show the pipeline stage graph.
This is due to /runs/N/nodes/ returning a 500 error, which is due to a null pointer exception.
And has been failing since the update to bismuth.
The stage view (classic) does not have this problem.
Still working on how exactly to reproduce this but thought the stack trace may be useful on its own.
Caused by: java.lang.NullPointerException at org.jenkinsci.plugins.workflow.actions.TimingAction.getStartTime(TimingAction.java:45) at org.jenkinsci.plugins.workflow.pipelinegraphanalysis.StatusAndTiming.computeChunkTiming(StatusAndTiming.java:240) at io.jenkins.blueocean.rest.impl.pipeline.PipelineNodeGraphVisitor.handleChunkDone(PipelineNodeGraphVisitor.java:201) at org.jenkinsci.plugins.workflow.graphanalysis.StandardChunkVisitor.chunkStart(StandardChunkVisitor.java:40) at io.jenkins.blueocean.rest.impl.pipeline.PipelineNodeGraphVisitor.chunkStart(PipelineNodeGraphVisitor.java:74) at org.jenkinsci.plugins.workflow.graphanalysis.ForkScanner.visitSimpleChunks(ForkScanner.java:566) at org.jenkinsci.plugins.workflow.graphanalysis.ForkScanner.visitSimpleChunks(ForkScanner.java:551) at io.jenkins.blueocean.rest.impl.pipeline.PipelineNodeGraphVisitor.<init>(PipelineNodeGraphVisitor.java:68) at io.jenkins.blueocean.rest.impl.pipeline.NodeGraphBuilder$NodeGraphBuilderFactory.getInstance(NodeGraphBuilder.java:37) at io.jenkins.blueocean.rest.impl.pipeline.PipelineNodeContainerImpl.<init>(PipelineNodeContainerImpl.java:32) at io.jenkins.blueocean.rest.impl.pipeline.PipelineRunImpl.getNodes(PipelineRunImpl.java:81) at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62) at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43) at java.lang.reflect.Method.invoke(Method.java:498) at org.kohsuke.stapler.Function$InstanceFunction.invoke(Function.java:324) at org.kohsuke.stapler.MetaClass$3.doDispatch(MetaClass.java:196) at org.kohsuke.stapler.NameBasedDispatcher.dispatch(NameBasedDispatcher.java:58) at org.kohsuke.stapler.Stapler.tryInvoke(Stapler.java:746)
Attachments
Issue Links
- is duplicated by
-
-
- Closed
-
Activity
OK did some digging.
The failure happens here:
https://github.com/jenkinsci/workflow-api-plugin/blob/master/src/main/java/org/jenkinsci/plugins/workflow/actions/TimingAction.java#L45
due to flowNode being null.
The ternary expression https://github.com/jenkinsci/pipeline-graph-analysis-plugin/blob/master/src/main/java/org/jenkinsci/plugins/workflow/pipelinegraphanalysis/StatusAndTiming.java#L240
does not guard against firstNode being null (I am not sure if the true/false is correct either, but can't comment on that). Clearly firstNode is null, and it is trying to find the timing for it.
The root cause seems to be that "firstExecuted" is null:
in PipelineNodeGraphVisitor.
vivek right there is a case where firstExecuted could be null, in fact there is a check for it:
TimingInfo times = StatusAndTiming.computeChunkTiming(run, chunk.getPauseTimeMillis(), firstExecuted, chunk.getLastNode(), chunk.getNodeAfter());
if(times == null){
times = new TimingInfo();
}
GenericStatus status = (firstExecuted == null) ? GenericStatus.NOT_EXECUTED :StatusAndTiming
.computeChunkStatus(run, chunk.getNodeBefore(), firstExecuted, chunk.getLastNode(), chunk.getNodeAfter());
However this check happens after it is attempted to be used to get the timing. It shouldn't be called if null.
Jenkins 2.28 + BlueOcean b11 - stack trace looks the same - but for the avoidance of doubt
As a test I created a multibranch pipeline test job that launches a secondary pipeline job (same structure as my failing job) - it did not have the same issue. So there is something subtle about it.