Backup container failing takes down database pod

If the pbm-agent sidecar for some reason cannot connect to MongoDB, the pod never gets to "ready" state, and CrashLoopBackOff ensues.

For example, if password is misconfigured in secrets this happens.

      name: backup-agent
      ready: false
      restartCount: 870
      started: false
      state:
        waiting:
          message: back-off 5m0s restarting failed container=backup-agent pod=percona-mongo-psmdb-d-rs1-1_data(f8278e2e-dc5f-4499-9de9-7c836d73a1a0)
          reason: CrashLoopBackOff

It might be better to have the database working, even if backups fail. Or perhaps current behavior is intentional?