Pre-Backup pod labels are too long

[SchoolGuy]

Description

The label of the pre-backup pod is too long. I have no clever way how to solve this but this should definitely not happen.

apiVersion: k8up.io/v1
kind: Schedule
metadata:
  name: backup-schedule-schoolguy-photoprism
  namespace: photoprism
spec:
  podSecurityContext:
    runAsUser: 0
  backend:
    rest:
      url: "http://<IP>:8000/<repository_name>"
    repoPasswordSecretRef:
      name: backup-repo
      key: password
  backup:
    schedule: '0 3 * * *'
    failedJobsHistoryLimit: 2
    successfulJobsHistoryLimit: 2
    promURL: http://prometheus-pushgateway.monitoring.svc.cluster.local:9091
  check:
    schedule: '0 5 * * 6'
    promURL: http://prometheus-pushgateway.monitoring.svc.cluster.local:9091
  prune:
    schedule: '0 0 1 * *'
    retention:
      keepLast: 5
      keepDaily: 14

Additional Context

No response

Logs

2023-05-25T19:35:44Z ERROR k8up.operator Reconciler error {"controller": "backup.k8up.io", "controllerGroup": "k8up.io", "controllerKind": "Backup", "Backup": {"name":"backup-schedule-schoolguy-photoprism-backup-j7s5z","namespace":"photoprism"}, "namespace": "photoprism", "name": "backup-schedule-schoolguy-photoprism-backup-j7s5z", "reconcileID": "313f0f45-4162-4af3-a009-584173651199", "error": "unable to createOrUpdate(\"backup-backup-schedule-schoolguy-photoprism-backup-j7s5z-prebackup\"): Job.batch \"backup-backup-schedule-schoolguy-photoprism-backup-j7s5z-prebackup\" is invalid: spec.template.labels: Invalid value: \"backup-backup-schedule-schoolguy-photoprism-backup-j7s5z-prebackup\": must be no more than 63 characters"}
sigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).reconcileHandler
 /home/runner/go/pkg/mod/sigs.k8s.io/controller-runtime@v0.13.1/pkg/internal/controller/controller.go:326
sigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).processNextWorkItem
 /home/runner/go/pkg/mod/sigs.k8s.io/controller-runtime@v0.13.1/pkg/internal/controller/controller.go:273
sigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).Start.func2.2
 /home/runner/go/pkg/mod/sigs.k8s.io/controller-runtime@v0.13.1/pkg/internal/controller/controller.go:234

Expected Behavior

Pre-Backup Pods spawn as expected.

Steps To Reproduce

1. Use schedule from above
2. Wait for error to appear in log

Version of K8up

2.7.1 (app) / 4.2.2 (helmchart)

Version of Kubernetes

v1.26.4+k3s1

Distribution of Kubernetes

k3s

[SchoolGuy]

I am deeply sorry that I report so many issues in recent days. I really love the software you build and I will continue to pay the early adopter tax. Before and including 2.7.0 everything was very easy and rock solid for me!

[Kidswiss]

@SchoolGuy

I appreciate you reportiing bugs, so we're aware of them and can fix them. :)

As a workaround here I'd suggest to shorten the name of your backup schedule.

[SchoolGuy]

@Kidswiss Ack. Thanks for the tip.

[marvinWolff]

Hey there,

just ran into this issue.

Are there any news regarding a fix for this?

Thank you!

[arska]

Confirmed — the label value k8up.io/preBackupPod: template.Name in operator/backupcontroller/prebackup_utils.go:80,91 exceeds the 63-character Kubernetes label value limit when the PreBackupPod name is long.

Same class of bug as:

• Snapshot ID truncation (fixed in Add fuzz testing with CI and fix snapshot ID panic #1165)
• Volume name truncation (bug: truncate volume names to 63 characters in BackupController executor #1092)

The fix would truncate template.Name to 63 characters when used as a label value, similar to the pattern in #1092.