1. Proposed title of this feature request Need option to suppress the PodDisruptionBudgetAtLimit for certain pods PDB managed by operators

What is the nature and description of the request? We need a way to suppress PodDisruptionBudgetAtLimit when, by design, some PDBs are single instances, and the related operator will address PDB issues while the node drains. For example, EDB PostgreSQL cluster will have only one primary and multiple SandBY, and the following PDB defined is accurate and can not have multiple Primary instances at a time.

$ oc get pdb
 
NAME MIN AVAILABLE MAX UNAVAILABLE ALLOWED DISRUPTIONS AGE
common-service-db-primary 1 N/A 0 158dkind: PodDisruptionBudget
spec:
minAvailable: 1
selector:
matchLabels:
 k8s.enterprisedb.io/cluster: common-service-db
 role: primary

EDB PostgreSQL operator monitors the node, and the pods status If a node is cordoned off, the operator detects and updates the pod on the cordoned node as StandBY and promotes a pod on different nodes as Primary. You can effectively remove the PDB primary containers for the node about to drain, and it will not be a problem.

There is no need for continues alerts getting generated draing otere real alerts . We can not change teh current alers and many cases were opened by cloud pak

Some discussion and proposal here from other products https://github.com/openshift/cluster-kube-controller-manager-operator/pull/333#issuecomment-572731383

3. Why does the customer need this? (List the business requirements here) Currtly the ffollowing alers getting generated and no option to sective supress the PDB alerts

Alertname Starts At Summary State PodDisruptionBudgetAtLimit 2024-09-09 14:49:07 UTC The pod disruption budget is preventing further disruption to pods. active

We had many IBM Cloud pak customers and OCP admins requesting to get the alerts fixed

4. List any affected packages or components.

All IBM CloudPak (CP) customers are affected. Many CP opertors use the EDB postgreSQL database for persistance. The problem offect all IBM Cloud pack deplyment on production server where admins and management are monitoering the alers 24 X7

All OCP customers using EDB Postgres for Kubernetes are affected https://catalog.redhat.com/software/container-stacks/detail/5fb41c88abd2a6f7dbe1b37b

nents.