Description of problem:

After setting tekton-dashboard-url to some non-empty value in config map pipelines-as-code and reverting back to empty string, the GitHub messages still use the old value until both PAC pods are recreated.

Prerequisites (if any, like setup, operators/versions):

OpenShift 4.12
Pipelines 1.10

Steps to Reproduce

1. oc edit cm pipelines-as-code -n openshift-pipelines
2. set tekton-dashboard-url (e.g. foo.example.com)
3. create a PR, check messages
4. set tekton-dashboard-url to empty string
5. type "/retest" on PR

Actual results:

Messages use the content of tekton-dashboard-url until pods are restarted

Expected results:

New setting is used

Reproducibility (Always/Intermittent/Only Once):

Build Details:

Additional info (Such as Logs, Screenshots, etc):