When a node is marked down for maintenance, such as with the node maintenance operator, KubeDaemonSetMisScheduled and KubeDaemonSetRolloutStuck alerts begin firing for all platform "openshift-*" daemonsets.

For a cluster with a lot of Red Hat operators, this can reach 20 or more alerts of each type (40 in total).

https://github.com/openshift/cluster-monitoring-operator/blob/release-4.18/assets/control-plane/prometheus-rule.yaml#L116

https://github.com/openshift/cluster-monitoring-operator/blob/release-4.18/assets/control-plane/prometheus-rule.yaml#L167

When a cluster is put into maintenance, the operator does add a new taint:
{"effect":"NoSchedule","key":"medik8s.io/drain"}

We should factor a node being in maintenance / cordoned / SchedulingDisabled when it is in Ready state for the purpose of alerting.

4.18.15

Always

    1. OpenShift 4.18 Cluster
    2. Install Node Maintenance Operator
    3. Put a node into maintenance:

apiVersion: nodemaintenance.medik8s.io/v1beta1
kind: NodeMaintenance
metadata:
  name: nodemaintenance-cr
spec:
  nodeName: worker-0.ocp418shared.tamlab.brq2.redhat.com
  reason: "node maint"

A lot of extra alerts that are not providing actionable information.