Skip to content

OpenEBS cleanup issue: wait_for_ready() hangs indefinitely when pod phase is Succeeded #176

Description

The wait_for_ready() method in https://github.com/microsoft/AIOpsLab/blob/main/aiopslab/service/kubectl.py#L103 does not account for the pods in terminal phases like Succeeded or Failed.
This causes timeouts when OpenEBS cleanup pods remain in the namespace.


Benchmark failure

  1. OpenEBS creates cleanup pods after PVC deletion
  2. These pods enter Succeeded phase but remain in namespace
  3. wait_for_ready("openebs") sees: 3 ready + 1 succeeded = 3/4 ready -> Hits the 300s timeout
  4. Every subsequent problem fails at self.kubectl.wait_for_ready("openebs")

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions