[JIRA] (JENKINS-43556) Stage View shows incorrect build result

I also see this with jenkins 2.73.3, pipeline stage view 2.9 and pipeline rest api plugin 2.9.

Build history view on the left shows success, stage view shows the whole row in red with no cell highlighted in stripey red.

The normal api for the build: <job>/<buildnum>/api/python?pretty=true says "result": "SUCCESS"

The wfapi: <job>/wfapi/run shows <buildnum> as having "status": "FAILED" . All stages within that report "SUCCESS".

The pipeline steps view shows SUCCESS for every step so it seems like a problem with the overall status from the wfapi.

This message was sent by Atlassian JIRA (v7.3.0#73011-sha1:3c73d0e)

russell.gallop@gmail.com (JIRA)

unread,

Feb 14, 2018, 6:34:02 AM2/14/18

to jenkinsc...@googlegroups.com

Russell Gallop assigned an issue to Sam Van Oort

Jenkins /

Change By:	Russell Gallop
Assignee:	Sam Van Oort

svanoort@cloudbees.com (JIRA)

unread,

Feb 14, 2018, 9:31:03 AM2/14/18

to jenkinsc...@googlegroups.com

Sam Van Oort assigned an issue to Unassigned

Jenkins /

Change By:	Sam Van Oort
Assignee:	Sam Van Oort

svanoort@cloudbees.com (JIRA)

unread,

Feb 14, 2018, 9:32:03 AM2/14/18

to jenkinsc...@googlegroups.com

Sam Van Oort commented on

Russell Gallop Adam Zovits Please include a job that will actually reproduce the issue – screenshots do not provide the info needed to debug – I only know that some job, somewhere had issues, and that's not very helpful.

adam.zovits@thalesgroup.com (JIRA)

unread,

Feb 15, 2018, 3:07:02 AM2/15/18

to jenkinsc...@googlegroups.com

Adam Zovits commented on

Sam Van Oort I'm sorry, but in my case this failure was never reliably reproducible. As shown on the screenshot, the next run without any changes to the code results in a green build. If that would help, I could provide an approximate plan of the jenkinsfile being used (the original jenkinsfile must never leave the company servers).

svanoort@cloudbees.com (JIRA)

unread,

Feb 15, 2018, 10:15:03 AM2/15/18

to jenkinsc...@googlegroups.com

Sam Van Oort commented on

Adam Zovits I would like to fix this, if we can find a way to reproduce it. If you could provide the pipeline structure, that helps some – if you could provide the status (pass/fail) of the final step and the build.xml (sanitized) that would help even more and could make this possible to fix.

adam.zovits@thalesgroup.com (JIRA)

unread,

Feb 15, 2018, 11:30:02 AM2/15/18

to jenkinsc...@googlegroups.com

Adam Zovits updated an issue

Jenkins /

Change By:	Adam Zovits
Attachment:	jenkinsfile_sanitized.txt

adam.zovits@thalesgroup.com (JIRA)

unread,

Feb 15, 2018, 11:34:03 AM2/15/18

to jenkinsc...@googlegroups.com

Adam Zovits commented on

Sam Van Oort I'm sorry, but we have no logs or build files from this long ago. I could only gather the jenkinsfile being used at that time, which I have attached in a sanitized version (removed all possible identifying names but none of the build steps that were actually executed): jenkinsfile_sanitized.txt

I hope that helps at least a bit.

russell.gallop@gmail.com (JIRA)

unread,

Feb 19, 2018, 7:52:02 AM2/19/18

to jenkinsc...@googlegroups.com

Russell Gallop edited a comment on

I also see this with jenkins 2.73.3, pipeline stage view 2.9 and pipeline rest api plugin 2.9.

Build history view on the left shows success, stage view shows the whole row in red with no cell highlighted in stripey red.

The normal api for the build: <job>/<buildnum>/api/python?pretty=true says "result": "SUCCESS"

The wfapi: <job>/wfapi/ run runs shows <buildnum> as having "status": "FAILED" . All stages within that report "SUCCESS".

The pipeline steps view shows SUCCESS for every step so it seems like a problem with the overall status from the wfapi.

russell.gallop@gmail.com (JIRA)

unread,

Feb 20, 2018, 11:08:05 AM2/20/18

to jenkinsc...@googlegroups.com

Russell Gallop updated an issue

Jenkins /

Change By:	Russell Gallop
Attachment:	JENKINS-43556.py

russell.gallop@gmail.com (JIRA)

unread,

Feb 20, 2018, 11:19:03 AM2/20/18

to jenkinsc...@googlegroups.com

Russell Gallop commented on

Similar to Adam, I don't think that it can be reproduced just from the pipeline DSL. I have seen this happen on one build and be okay on the ones just before and after.

I have produced a small script to find instances of this problem by spotting differences between the wfapi and the main Jenkins API (attached) JENKINS-43556.py. This didn't find any more examples of the problem so it seems that it happened 3 times within about 24 hours but not since.

We have since restarted Jenkins (and updated 2.73.3 -> 2.98.3) and can no longer see the problem with the previous builds in the wfapi, they have long since disappeared from the pipeline view.

I had a look at the workflow files in the Jenkins filesystem for these jobs (after the restart) and can't see any problems (e.g. incorrect status) in the files there.

JENKINS-43556.py

russell.gallop@gmail.com (JIRA)

unread,

Feb 26, 2018, 11:33:03 AM2/26/18

to jenkinsc...@googlegroups.com

Russell Gallop commented on

I've seen this again and checked the files in the file store. I can't see any problems in them. I don't think I can send them all but I have a copy stored so if there is anything useful for me to look for I can.

adam.zovits@thalesgroup.com (JIRA)

unread,

Mar 14, 2018, 12:08:03 PM3/14/18

to jenkinsc...@googlegroups.com

Adam Zovits commented on

I too have seen this issue for the second time and have made a backup copy of the workspace and the job directory in case any of the files there can be of help to analyse the error.

c.amshoff@gmx.de (JIRA)

unread,

Mar 16, 2018, 10:00:02 AM3/16/18

to jenkinsc...@googlegroups.com

Christoph Amshoff commented on

Same issue here for maybe one out of 30 builds; up to now this seems to happen just for the first successful build after a broken (failed) build. Build history is fine. logs don't show any issue, just the pipeline in stage view is rendered red. In Blue Ocean, everything looks good.

I don't know if that heps, but when looking at the rendered HTML of stage view, I can see that tr has CSS classes "job FAILED", while all td's for the steps have classes "stage-cell stage-cell-x SUCCESS".

We're using Jenkins 2.89.4, Pipeline Stage View Plugin 2.9 and Pipeline REST API plugin 2.9.

michele.vanini@gmail.com (JIRA)

unread,

Jun 19, 2018, 8:37:02 AM6/19/18

to jenkinsc...@googlegroups.com

Michele Vanini commented on

Same issue on CloudBees Jenkins Enterprise 2.121.1.2-rolling with Stage View Plugin 2.9

mhare@zixcorp.com (JIRA)

unread,

Jun 25, 2018, 5:43:01 PM6/25/18

to jenkinsc...@googlegroups.com

Michael Hare commented on

I have this problem as well. I think I caused it by deleting all my jobs and resetting my next build number to 1. I had been testing some pipeline stuff and had run up 75 or so useless builds. I used the UI, deleted each one, the reset the next build number to 1. It works, except the Stage View is wrong. I think it might be showing the results (e.g. Job 22) for the first time I had a "Job 22"

This message was sent by Atlassian JIRA (v7.10.1#710002-sha1:6efc396)

mhare@zixcorp.com (JIRA)

unread,

Jun 25, 2018, 5:53:03 PM6/25/18

to jenkinsc...@googlegroups.com

Michael Hare edited a comment on

I have this problem as well. I think I caused it by deleting all my jobs and resetting my next build number to 1. I had been testing some pipeline stuff and had run up 75 or so useless builds. I used the UI, deleted each one, the then reset the next build number to 1. It works, except the Stage View is wrong. I think it might be showing the results (e.g. Job 22) for the first time I had a "Job 22"

russell.gallop@gmail.com (JIRA)

unread,

Jun 28, 2018, 4:46:01 AM6/28/18

to jenkinsc...@googlegroups.com

Russell Gallop commented on

Michael Hare that may be the same symptom but I can confirm that we did not do anything like that to cause this issue. This is coming up on fresh builds with numbers which hadn't been used before.

michele.vanini@gmail.com (JIRA)

unread,

Aug 9, 2018, 11:27:01 AM8/9/18

to jenkinsc...@googlegroups.com

Michele Vanini commented on

In my case it seems related to JEP-200: some plugins weren't updated caused Beekeeper/CAP policy.
If jenkins was unable to serialize some object for write in xml, it fails the build in a strange manner that result in this issue: history item green and stage-view red without failed stage.

eslam.husseiny@gmail.com (JIRA)

unread,

Aug 28, 2018, 9:48:02 AM8/28/18

to jenkinsc...@googlegroups.com

Eslam ElHusseiny commented on

Same issue with Jenkins v2.135, and Pipeline Stage View Plugin 2.10.

This message was sent by Atlassian Jira (v7.11.2#711002-sha1:fdc329d)

gunter@grodotzki.co.za (JIRA)

unread,

Oct 16, 2018, 4:43:03 AM10/16/18

to jenkinsc...@googlegroups.com

Gunter Grodotzki commented on

Same issue with Jenkins 2.138.2 + Pipeline Stage View Plugin 2.10

joeywazen@gmail.com (JIRA)

unread,

Oct 19, 2018, 9:02:02 AM10/19/18

to jenkinsc...@googlegroups.com

Joey Wazen updated an issue

Jenkins /

Change By:	Joey Wazen
Priority:	Minor Major

svanoort@cloudbees.com (JIRA)

unread,

Oct 19, 2018, 9:09:02 AM10/19/18

to jenkinsc...@googlegroups.com

Sam Van Oort updated an issue

Jenkins /

Change By:	Sam Van Oort
Priority:	Major Minor

joeywazen@gmail.com (JIRA)

unread,

Oct 23, 2018, 5:56:03 AM10/23/18

to jenkinsc...@googlegroups.com

Joey Wazen assigned an issue to Sam Van Oort

Jenkins /

Change By:	Joey Wazen
Assignee:	Sam Van Oort

phil.segal@gmail.com (JIRA)

unread,

Nov 20, 2018, 7:02:03 AM11/20/18

to jenkinsc...@googlegroups.com

Phil Segal updated an issue

Jenkins /

Change By:	Phil Segal
Priority:	Minor Major

phil.segal@gmail.com (JIRA)

unread,

Nov 20, 2018, 7:02:08 AM11/20/18

to jenkinsc...@googlegroups.com

Phil Segal assigned an issue to Phil Segal

Jenkins /

Change By:	Phil Segal
Assignee:	Sam Van Oort Phil Segal

phil.segal@gmail.com (JIRA)

unread,

Nov 20, 2018, 7:03:06 AM11/20/18

to jenkinsc...@googlegroups.com

Phil Segal assigned an issue to Sam Van Oort

Jenkins /

Change By:	Phil Segal
Assignee:	Phil Segal Sam Van Oort

phil.segal@gmail.com (JIRA)

unread,

Nov 20, 2018, 7:04:02 AM11/20/18

to jenkinsc...@googlegroups.com

Phil Segal commented on

This is a fairly fundamental issue, it should surely be more than a minor issue. Causes a lot of confusion and distrust.

majgis@gmail.com (JIRA)

unread,

Dec 4, 2018, 6:49:02 PM12/4/18

to jenkinsc...@googlegroups.com

Michael Jackson commented on

We were just hit by this issue again in 2.10

aburdajewicz@cloudbees.com (JIRA)

unread,

Dec 24, 2018, 12:52:02 AM12/24/18

to jenkinsc...@googlegroups.com

Allan BURDAJEWICZ commented on

I have seen cases for which a restart solved this. Anybody able to workaround this by a restart ?

aburdajewicz@cloudbees.com (JIRA)

unread,

Dec 26, 2018, 7:45:06 PM12/26/18

to jenkinsc...@googlegroups.com

Allan BURDAJEWICZ edited a comment on

I have seen cases for which a restart solved this. Anybody able to workaround this by a restart ?

I wonder if this could be due to the caching of some objects, for example [com.cloudbees.workflow.flownode.FlowNodeUtil|https://github.com/jenkinsci/pipeline-stage-view-plugin/blob/pipeline-stage-view-2.10/rest-api/src/main/java/com/cloudbees/workflow/flownode/FlowNodeUtil.java] or [js.model.rest-api.js|https://github.com/jenkinsci/pipeline-stage-view-plugin/blob/pipeline-stage-view-2.10/ui/src/main/js/model/rest-api.js]

luke.southwell@ultrasoc.com (JIRA)

unread,

Feb 27, 2019, 9:57:02 AM2/27/19

to jenkinsc...@googlegroups.com

Luke Southwell commented on

I am also seeing this issue which has caused great confusion. Has anyone got a reliable work around?

luke.southwell@ultrasoc.com (JIRA)

unread,

Feb 27, 2019, 12:53:03 PM2/27/19

to jenkinsc...@googlegroups.com

Luke Southwell edited a comment on

I am also seeing this issue with 2.150.3 which has caused great confusion. Has anyone got a reliable work around?

cwagner118@gmail.com (JIRA)

unread,

Apr 10, 2019, 4:47:02 PM4/10/19

to jenkinsc...@googlegroups.com

Chris Wagner commented on

This seems to manifest itself when build numbers are re-used. Consider the following scenario:

Builds 1-10 all pass, then builds 11-20 all fail, you decide to clean that up once you fix the build. You delete 11-20 and set your next build number to 11. Now the stages view reports failures for 11-20 even though they are now actually successful.

miles@genasys.co.za (JIRA)

unread,

Apr 15, 2019, 4:18:02 AM4/15/19

to jenkinsc...@googlegroups.com

Miles Gray Buckton commented on

I'm also experiencing this issue randomly. It'll be great if it can be prioritized, and fixed.

stanley.ky.lee@gmail.com (JIRA)

unread,

May 27, 2019, 6:57:03 PM5/27/19

to jenkinsc...@googlegroups.com

Stanley Lee commented on

Allan BURDAJEWICZ, was able to work around this by doing a restart. I'm suspecting some sort of caching as well. I was running 2.169 and updated to 2.179 which restarted the server. The correct UI visualizations came up after the restart.

stanley.ky.lee@gmail.com (JIRA)

unread,

May 27, 2019, 7:11:06 PM5/27/19

to jenkinsc...@googlegroups.com

Stanley Lee edited a comment on

[~allan_burdajewicz], I was able to work around this by doing a restart. I'm suspecting some sort of caching as well. I was running 2.169 and updated to 2.179 which restarted the server. The correct UI visualizations came up after the restart.

raihaan.shouhell@autodesk.com (JIRA)

unread,

Oct 31, 2019, 12:11:06 AM10/31/19

to jenkinsc...@googlegroups.com

Raihaan Shouhell commented on

Allan BURDAJEWICZ I am experiencing this issue and when checking the cache. Indeed the FlowNodeUtil cache RunExt is a failure instead of success

This message was sent by Atlassian Jira (v7.13.6#713006-sha1:cc4451f)

hannes.strautmann@metrilus.de (JIRA)

unread,

Nov 27, 2019, 8:03:06 AM11/27/19

to jenkinsc...@googlegroups.com

Hannes Strautmann commented on

We are also experiencing this issue lately, randomly. We have not deleted any builds nor reset the next build number.

Jenkins LTS 2.190.2, all plugins "up to date" (not older than say 4 weeks).

Same observations as Russell Gallop:

The normal api for the build: <job>/<buildnum>/api/python?pretty=true says "result": "SUCCESS"
The wfapi: <job>/wfapi/runs shows <buildnum> as having "status": "FAILED" . All stages within that report "SUCCESS".

marat@slonopotamus.org (JIRA)

unread,

Dec 25, 2019, 7:43:04 AM12/25/19

to jenkinsc...@googlegroups.com

Marat Radchenko commented on

Confirming what Hannes Strautmann said in previous comment. I also observe situation where wfapi/runs reports FAILED for build that is actually successful.

marat@slonopotamus.org (JIRA)

unread,

Dec 26, 2019, 3:16:03 AM12/26/19

to jenkinsc...@googlegroups.com

Marat Radchenko commented on

Note that PR#75 was merged. While it doesn't fix the root issue (invalid data getting into RunExt cache), it will catch and invalidate bogus cached data when it is got from cache.

This change wasn't yet released in any pipeline-stage-view-plugin version.

Kalle.Niemitalo@procomp.fi (JIRA)

unread,

Jan 22, 2020, 5:11:03 AM1/22/20

to jenkinsc...@googlegroups.com

Kalle Niemitalo commented on

I was able to remove the incorrect data from the cache by calling com.cloudbees.workflow.flownode.FlowNodeUtil.CacheExtension.all().get(0).getRunCache().invalidate("the externalizable ID of the run") from the script console.

Before that, I had tried "Reload Configuration from Disk" under "Manage Jenkins", but it had not cleared the cache, even though its description appears to say otherwise: "Discard all the loaded data in memory and reload everything from file system. Useful when you modified config files directly on disk."

raihaan.shouhell@autodesk.com (JIRA)

unread,

Jan 31, 2020, 4:34:10 AM1/31/20

to jenkinsc...@googlegroups.com

Raihaan Shouhell resolved as Fixed

Should be resolved please reopen if you encounter this issue again

Jenkins /

Change By:	Raihaan Shouhell
Status:	Open Resolved
Assignee:	Sam Van Oort Raihaan Shouhell
Resolution:	Fixed
Released As:	pipeline-stage-view 2.13