Pod disruption budget: tekton-pipelines-webhook issue

[bitsofinfo]

Hi, experienced this w/ 0.18.1 on Azure AKS during an rolling upgrade of k8s on the cluster via AKS's platform.

I was having troubles not being able to upgrade any cluster; AKS provisions a new node, drains one then upgrades and on and on. The entire process was hanging and failing, and it turned out due to the PodDisruptionBudget for tekton-pipelines-webhook. Is this expected? What should be the recommended procedure for avoiding this situation? Deleting the PDB prior to these kind of operations?

Name:           tekton-pipelines-webhook
Namespace:      tekton-pipelines
Min available:  80%
Selector:       app.kubernetes.io/component=webhook,app.kubernetes.io/instance=default,app.kubernetes.io/name=webhook,app.kubernetes.io/part-of=tekton-pipelines
Status:
    Allowed disruptions:  0
    Current:              1
    Desired:              1
    Total:                1
Events:                   <none>

[bitsofinfo]

or i guess just scaling it prior to an upgrade would solve it

[imjasonh]

Duplicate of #3654 -- I'll comment over there