backup stuck in "requested" status after it started while another backup was running

[maslosan]

Report

Scheduled backup stuck in "requested" status, following backups stuck in "waiting" status

More about the problem

I'm started manual backup at 14:42 (backup needs ~8 hours to complete) by the CR (kubectl apply -f backup.yaml):

apiVersion: psmdb.percona.com/v1
kind: PerconaServerMongoDBBackup
metadata:
  finalizers:
  - percona.com/delete-backup
  labels:
    ancestor: daily
    cluster: mongo-db-common
    type: cron
  name: cron-mongo-db-c-test
  namespace: mongo-db
spec:
  clusterName: mongo-db-common
  compressionType: zstd
  storageName: minio
  type: logical

after 15 minutes i deleted this CR (kubectl delete -f backup.yaml), and I think what backup-agent did not cancel this backup and scheduled backup (cron at 20:45) start at a time, when backup did not complete yet, and stuck in "requested" state, and follow scheduled backup stuck in "waiting" state

Steps to reproduce

1. Create cluster mongodb with scheduled backup
2. Create and run PerconaServerMongoDBBackup test.yaml with starting backup before running a scheduled backup, in less time than it takes to create this backup
3. Delete PerconaServerMongoDBBackup test.yaml manifest when it in status "Running"
4. Check new scheduled backup status

Versions

1. Kubernetes v1.30.13
2. Operator v1.17.0
3. Database percona/percona-server-mongodb:6.0.15-12

Anything else?

No response

[gkech]

@nmarukovich can you try to reproduce please?

[nmarukovich]

I can’t reproduce the issue on the main branch. The demand backup is deleted successfully, and the scheduled backup runs and completes as expected.