Details
-
Improvement
-
Status: In Progress (View Workflow)
-
Major
-
Resolution: Unresolved
Description
When a flow involves many steps, long logs, and/or many branches, it can be hard for a developer receiving a failure email (for example) to quickly see what part of the build actually failed and why. I think it would be useful for the catchError step to do a DFS search on the Throwable and its cause chain through the FlowNode graph starting at the end of the catch step, looking for ErrorAction, and setting/appending an environment variable with the URL of the LogAction. Or provide a new step to do the same. Thus you could write
def runStuff(param) {
...
}
try {
parallel a: {runStuff 'a'}, b: {runStuff 'b'}
} catch (e) {
mail to: '...', subject: 'Failure!', body: "Build failed: ${errorUrl(e)}"
}
and get a link to http://jenkins/job/myflow/123/flowGraph/77/console or the like, according to the actual step in one of the branches that threw the error.
Attachments
Issue Links
- depends on
-
-
- Open
-
- is duplicated by
-
JENKINS-38156 Need branch name and other details from parallel branches
-
- Resolved
-
- is related to
-
JENKINS-25894 Ugly reporting of errors from parallel step
-
- Resolved
-
-
JENKINS-26107 Allow stage to operate as a labelled block
-
- Closed
-
- relates to
-
-
- Resolved
-
- links to
Activity
erikvv if you are using blue ocean, you can just have the link to the build and once the page open it will automatically scroll to the failed step.
also another work around https://stackoverflow.com/questions/53444196/get-log-for-each-jenkins-pipeline-stage
thang bui
added a comment - - edited erikvv if you are using blue ocean, you can just have the link to the build and once the page open it will automatically scroll to the failed step.
also another work around https://stackoverflow.com/questions/53444196/get-log-for-each-jenkins-pipeline-stage 
Edgars Batna
added a comment - +1 The advent of Blue Ocean came and went. Jenkins is still a powerful tool, but I feel weird seeing many functional improvements get so stale.
Edgars Batna
added a comment - +1 The advent of Blue Ocean came and went. Jenkins is still a powerful tool, but I feel weird seeing many functional improvements get so stale. Late in the game, but this is still not straightforward as it could be.
Is there already any way to get the link to 'blue/rest/organizations/jenkins/pipelines/<jobname>/runs/<build-id>/nodes/<PipelineNodeImpl-id>/log/ from within a pipeline parallel node?
Apparently it's possible but, unfortunately, not out-of-the-box:
- For getting the Pipeline implementation node ID, take a look at Unexpected exception when use getContext and sh together in the latest version of Jenkins. The example even works from non-library code!
- For building a link , check out Jenkins: Generating Blue Ocean URLs. The example even offers to choose between legacy or BlueOcean!
Seems that getting the link (out-of-the-box) is now closer with the RUN_DISPLAY_URL environment variable injected by the BlueOcean plugin but it would be very nice to have an environment variable set to the step ID available throughout the whole job (possibly as part of the pipeline plug-in?). Not only for linking but also for obtaining a unique identifier which maps to something and can be used for (temporary?) storage path, artifact filenames, for better diagnostic output, etc.
Edit: correct link for "building a link"; added some arguments.
+1 for this.
Currently we have to perform some tricks to print URL to the necessary BO pipeline node from the groovy job itself.
First, RUN_DISPLAY_URL is not convenient if you want to add smth like "/pipeline/<node_id>" to it. The thing is that this var just adds /display/redirect to the classic URL, so we get smth like https://<jenkins_host>/<job_name>/<build_id>/display/redirect which is then redirected to https://<jenkins_host>/blue/organizations/jenkins/<job_name>/<build_id>/pipeline
So we first need to now the final URL (after redirections) - one can do this for example using [https://gist.github.com/sbglasius/3899754|this snippet]
Next, we should find necessary node id for the current stage. This step looks ugly, we just walk through the flow graph looking for a node with a needed name. It works for our pipelines consisting of many parallel jobs, suggestions are welcome for better and more universal code:
Integer getFlowID() {
FlowGraphWalker walker = new FlowGraphWalker(currentBuild.rawBuild.getExecution())
for (FlowNode flowNode: walker) {
if (flowNode.getDisplayName() == "Branch: " + env.STAGE_NAME) {
return flowNode.getId()
}
}
return 0
}
Definitely having a single variable for this would be a great solution.
Denis
added a comment - - edited +1 for this.
Currently we have to perform some tricks to print URL to the necessary BO pipeline node from the groovy job itself.
First, RUN_DISPLAY_URL is not convenient if you want to add smth like "/pipeline/<node_id>" to it. The thing is that this var just adds /display/redirect to the classic URL, so we get smth like https://<jenkins_host>/<job_name>/<build_id>/display/redirect which is then redirected to https://<jenkins_host>/blue/organizations/jenkins/<job_name>/<build_id>/pipeline
So we first need to now the final URL (after redirections) - one can do this for example using [https://gist.github.com/sbglasius/3899754|this snippet]
Next, we should find necessary node id for the current stage. This step looks ugly, we just walk through the flow graph looking for a node with a needed name. It works for our pipelines consisting of many parallel jobs, suggestions are welcome for better and more universal code:
Integer getFlowID() {
FlowGraphWalker walker = new FlowGraphWalker(currentBuild.rawBuild.getExecution())
for (FlowNode flowNode: walker) {
if (flowNode.getDisplayName() == "Branch: " + env.STAGE_NAME) {
return flowNode.getId()
}
}
return 0
}
Definitely having a single variable for this would be a great solution.
Perhaps a workaround would be to pipe the output within the step to a file and publish it as an artifact.