Jenkins doesn't have label XYZ

[ppitonak]

Issue Overview

Pipeline gets stuck at the beginning and Jenkins log only shows message 'Jenkins' doesn't have label 'mavenPROJECTNAME_1' and a spinner

Expected Behaviour

1. Jenkins slave is created and pipeline finishes successfully
2. The message explains the situation more clearly

Current Behaviour

Steps To Reproduce

1. create new space
2. create new SpringBoot app
3. open pipeline log

Additional Information

Observable in this job https://ci.centos.org/job/devtools-test-e2e-openshift.io-smoketest-us-east-2-released/1506/console

The broken functionality is a regression, originally solved in #4020

The UX part was supposed to be fixed by #3797

[chmouel]

The broken functionality is a regression, originally solved in #4020

I don't believe this is solved in 4020,

"Jenkins doesn't have label" is a catch all sentence, which in english would mean, "i have asked to get a new pod with this label but it didn't came up"

I can't get a pod can mean a lot of things, 99% of the time is beause of an infrastrcture issue, without logs it's hard to say,

[piyush-garg]

@ppitonak If the previous slave does not get deleted by any reason, then Jenkins sometimes not able to start new slave pod and that causes this. And it is the message coming from Jenkins, not from any of our component, we can try to make this message more explanatory but can not assure you anything.

[ppitonak]

@chmouel ok, makes sense...time for your team to dive into logs :)

@piyush-garg the message was already changed in #3797. Why do I see the old one again?

[piyush-garg]

@ppitonak The changed message appears when slave pod gets ready, but till that time it does not print anything. In our case, the slave pod was not provisioned that's why we did not see the message, otherwise, if you see any other job this message is still there.

[jarifibrahim]

This could be related to #4745. #4745 was also seen on cluster 2.

[Preeticp]

@jarifibrahim you are right, as mentioned in the last screenshot for #4745 I get a similar error message in my second attempt and the pipeline seems to be stuck forever (19 hours now).