[ws-manager] enable the volume snapshot controller when the VolumeSnapshot CRD exists

[jenting]

Description

For self-hosted, the VolumeSnapshot is optional for now, we should enable the volume_snapshot_controller only when the VolumeSnapshot CRD exists.

Related Issue(s)

Related to #10887
https://gitpod.slack.com/archives/C01KLC56NP7/p1656394622810719?thread_ts=1656366685.819199&cid=C01KLC56NP7

How to test

1. Scale down the volume snapshot controller.

   kubectl scale --replicas=0 deploy/snapshot-controller -n kube-system

2. Delete the VolumeSnapshot/VolumeSnapshotContent/VolumeSnapshotClass CRDs.

   kubectl delete vsclass --all
   kubectl delete crds volumesnapshotclasses.snapshot.storage.k8s.io
   kubectl delete crds volumesnapshots.snapshot.storage.k8s.io
   kubectl delete crds volumesnapshotcontents.snapshot.storage.k8s.io

3. Restart the ws-manager, and make sure the ws-manager is able to start.

   kubectl rollout restart deploy/ws-manager
   kubectl rollout status deploy/ws-manager
   kubectl get pod -l component=ws-manager

Release Notes

[experimental] enable the volume snapshot controller when the VolumeSnapshot CRD exists

Documentation

N/A

Werft options:

• /werft with-preview

[jenting]

/hold

[jenting]

/unhold

[Pothulapati]

/werft run publish-to-kots

👍 started the job as gitpod-build-jenting-ws-manager-check-volumesnapshotcrd.2
(with .werft/ from main)

[jenting]

/werft run publish-to-kots with-clean-slate-deployment

👍 started the job as gitpod-build-jenting-ws-manager-check-volumesnapshotcrd.3
(with .werft/ from main)

[mrsimonemms]

Confirm that this works in the June release on a self-hosted cluster