Description:
Operator workload-deployment runs on Kubernetes cluster without problem.
Afterwards StatefulSet is created for mongodb-cluster to manage 3 pods.
Pods crash with crashloopbackoff error.
value.yaml under psmdb-db folder looks like this:
replsets:
rs0:
podSecurityContext:
fsGroup: 9001
supplementalGroups:
- 1001
containerSecurityContext:
allowPrivilegeEscalation: false
capabilities:
drop:
- ALL
privileged: false
readOnlyRootFilesystem: true
runAsGroup: 9001
runAsNonRoot: true
runAsUser: 9001
Steps to Reproduce:
-
cd charts/psmdb-operator
-
helm upgrade
–install mongodb-operator-1
–debug
–wait
–timeout=20m . -
cd charts/psmdb-db
-
helm upgrade
–install mongodb-cluster-1
–debug
–set serviceAccount.name=percona-server-mongodb-operator
–wait
–timeout=20m .
Version:
1.20.0
Logs:
Logs from Pod.
path":“/tmp/mongodb-27017.sock”,“error”:"Read-only file system