Idler intermittently stopping active build pipelines on starter-us-east-2

[ldimaggi]

Issue Overview

Multiple instances of the idler NOT idling/un-idling when expected were seen on January 7. The behavior is not being seen on January 8. Logging this issue to document the behavior.

Expected Behaviour

Current Behaviour

The idler was not able to un-idle when a new build pipeline was created. Also, the idler was triggered when a new build pipeline was created. This behavior was seen repeatedly and consistently on January 7.

Steps To Reproduce

1. Perform an environment reset in OpenShift.io
2. Verify in the OpenShift console that the Jenkins pod has successfully started
3. Wait anywhere from several minutes to several hours
4. Observe that despite no build pipelines being created, the Jenkins pod is not idled.
5. Create a new app
6. Observe that when the build pipeline is created, the Jenkins pod is idled

[ldimaggi]

Seeing this today too - intermittently.

[ldimaggi]

The user account where this error is seen is provisioned on: starter-us-east-2
The booster used was Vert/X/REST HTTP

[piyush-garg]

We fixed #4739 issue on cluster 2 w.r.t to Idler. token was expired there. Can you please check if it's happening and if not then can you please close. Thanks