k8snetworkplumbingwg/whereabouts

[BUG] stuck in a leader election loop

caribbeantiger opened this issue · 3 comments

After reboot, a delete pod event causes whereabouts to get stuck in a leader election loop, and pod never goes out of this state.

Steps to reproduce the behavior:

  1. Reboot the worker node (might need to do it several times)
  2. delete pod
  3. pod stuck deleting since no response from whereabouts.
  4. get whereabout logs and see error : error retrieving resource lock /whereabouts: an empty namespace may not be set when a resource name is provided

Environment:

  • Whereabouts version : v0.6
  • Others: AWS EKS 1.24

Additional info / context
Add any other information / context about the problem here.
whereabouts logs.txt
whereabouts pod describe.txt

any luck looking into this?