Relax probing for forbidden resources

[hoerup]

Useing latest sloop on a 1.20.x cluster

In the log it seems that sloop continues to query for resources that it doesn't have access to

IMHO, when sloop has detected a forbidden - it should mark that resourcetype as not accessible and back off
(it could then perhaps retry at a later point eg every CrdRefreshInterval * 4 )

E0328 10:40:10.785774 1 reflector.go:156] pkg/mod/k8s.io/[email protected]/tools/cache/reflector.go:108: Failed to list *unstructured.Unstructured: servicemonitors.monitoring.coreos.com is forbidden: User "system:serviceaccount:sloop:sloop" cannot list resource "servicemonitors" in API group "monitoring.coreos.com" at the cluster scope
I0328 10:40:10.986713 1 trace.go:116] Trace[320645588]: "Reflector ListAndWatch" name:pkg/mod/k8s.io/[email protected]/tools/cache/reflector.go:108 (started: 2022-03-28 10:39:58.587890641 +0000 UTC m=+1821.020996664) (total time: 12.398775147s):

$kubectl logs --since=30m -n sloop sloop-0 | grep servicemonitors | wc -l
134
so it queries for the same forbidden resource type aproximately 4.47 times pr minute

[hoerup]

Or maybe event better: add config options for positivelist or negativelist resourcegroups to monitor