We had a build complete successfully; it shows green when viewing the details for that particular build and in the list of builds for that job. However, in the stage view for the job, that particular build shows red. Screenshot attached.
I also had this happen to me in prod twice in the 24 hours since I upgraded to latest started happening after we took in the Jenkins LTS.
This has happened with both regular pipeline and multi branch we upgraded pipeline-stage-tags-metadata 1.3.2 pipeline-stage-view 2.10 On Smithers Staging pipeline-stage-tags-metadata 1.3.8 On Smithers Staging pipeline-stage-view 2.11 among many other plugins, though those seem most applicable We also moved from 2.89.3 LTS to 2.164.3LTS
I also had this happen to me in prod twice in the 24 hours since I upgraded to latest started happening after we took in the Jenkins LTS.
This has happened with both regular pipeline and multi branch we upgraded pipeline-stage-tags-metadata 1.3.2 pipeline-stage-view 2.10 On Smithers Staging pipeline-stage-tags-metadata 1.3.8 On Smithers Staging pipeline-stage-view 2.11 among many other plugins, though those seem most applicable We also moved from 2.89.3 LTS to 2.164.3LTS
!Screen Shot 2019-06-03 at 1.27.23 PM.png! PR pipelines
I also had this happen reported to me in prod twice in the 24 hours since I upgraded to latest started happening after we took in the Jenkins LTS.
This has happened with both regular pipeline and multi branch we upgraded pipeline-stage-tags-metadata 1.3.2 pipeline-stage-view 2.10 On Smithers Staging pipeline-stage-tags-metadata 1.3.8 On Smithers Staging pipeline-stage-view 2.11 among many other plugins, though those seem most applicable We also moved from 2.89.3 LTS to 2.164.3LTS
We have noticed the same on our Multibranch pipelines. All stages are good, but Stage View still shows the build as failed. It is very rare when it occurs.
A user ran into the same problem. I did some quick debugging and it the problem is not on the client side. The rest API call incorrectly reports back failure. On the server side I found that calling RunExt.createNew() provides a RunExt with correct status, while RunExt.create() returns a cached RunExt with its status set to failure. I have no idea how that RunExt with incorrect status ended up in the cache. Would be interesting to know it it was had failure status when initially created or had success as initial result but the field was altered by some other code. No one knows...