karmada
karmada copied to clipboard
karmada-io
work may failed to create k8s resource due to label value over length limit
What happened:
The work created k8s resource, comes with a work.karmada.io/name label, which value is generated by this method, during mergeLabel, for back referencing work by k8s resource.
However, a name of a k8s object has a limit of length 253, it's not likely someone will create such a long named object, since it's really hard to remember, and hard to type in. But, limit of label values is far shorter than limit of name length, which is 63, and labels sometimes is generated by program, for referencing, not for typing or naming, it's easy to reach the limit.
And this happend in karmada.
https://github.com/karmada-io/karmada/blob/d27faeac953043c2274bb8787bbbaaafa7d0dcd6/pkg/util/names/names.go#L77-L99
What you expected to happen:
Since name of k8s object didn't reach the limit of name (253), karmada should succefully propagate it to member cluster.
How to reproduce it (as minimally and precisely as possible):
Propatate a k8s resource to member cluster with a name longer than 52 (plus -0123456789 suffix will reach the 63 limit)
Anything else we need to know?:
Environment:
- Karmada version:
- kubectl-karmada or karmadactl version (the result of
kubectl-karmada versionorkarmadactl version): - Others:
func mergeLabel(workload *unstructured.Unstructured, workNamespace string, binding metav1.Object, scope apiextensionsv1.ResourceScope) map[string]string {
var workLabel = make(map[string]string)
util.MergeLabel(workload, workv1alpha1.WorkNamespaceLabel, workNamespace)
util.MergeLabel(workload, workv1alpha1.WorkNameLabel, names.GenerateWorkName(workload.GetKind(), workload.GetName(), workload.GetNamespace()))
util.MergeLabel(workload, util.ManagedByKarmadaLabel, util.ManagedByKarmadaLabelValue)
...
return workLabel
}
Indeed
It's a kind of known issue. see https://github.com/karmada-io/karmada/issues/1741.
My idea is to move labels to annotation, but we need to analyze how these labels are used by the system and work out a backward-compatible solution.
@XiShanYongYe-Chang: Closing this issue.
In response to this:
It has fixed by #4865 /close
Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository.
