Bug: ServiceImport end­points not updated when re‑using an application namespace without cleanup

[Sumon009838]

📜 Description

When offboarding a namespace from a Slice and then deleting that Slice. Suppose I immediately create a new Slice that re‑uses the same application namespace (without first deleting the namespace). In that case, the new ServiceExport and ServiceImport appear to be created, but ServiceImport .status.endpoints is empty. Deleting the namespace before re-onboarding it fixes the problem, but users shouldn’t have to manually prune the namespace in between Slice lifecycles.

👟 Reproduction steps

1. Create a Slice (slice-a) that includes application namespace: app-ns on worker cluster(s).

2. Verify ServiceExport and ServiceImport status.endpoints are populated.

3. Offboard app-ns and delete slice-a.

4. Do not delete the app-ns namespace itself.

5. Create a new Slice (slice-b) that again includes the application namespace: app-ns.

6. Observe that ServiceImport CR are present in worker clusters, but their status. endpoints array is empty.

👍 Expected behavior

When a namespace is offboarded and then re‑onboarded without namespace deletion, the worker operator should:

1. Reconcile endpoints for the newly created ServiceExport so that .status.endpoints are correctly set.

👎 Actual Behavior

1. New ServiceImport is created, but their .status.endpoints remain empty.

2. The only workaround is to delete the entire namespace and recreate it.

🐚 Relevant log output

Version

No response

🖥️ What operating system are you seeing the problem on?

No response

✅ Proposed Solution

No response

👀 Have you spent some time to check if this issue has been raised before?

• I checked and didn't find any similar issue

Code of Conduct

• I agree to follow this project's Code of Conduct