Uploaded image for project: 'Couchbase Kubernetes'
  1. Couchbase Kubernetes
  2. K8S-542

CB instances never recover due to pod being alive

    XMLWordPrintable

Details

    • Bug
    • Resolution: Fixed
    • Critical
    • 1.0.0
    • 1.0.0
    • operator
    • None

    Description

      In TestPersistentVolumeNodeFailover, we create 6 node cluster with pvc, kill pod 0001 and 0005, wait for recovery, then crash nsserver on both recovered pods by moving files around inside those pods, then await recovery of both pods. This recovery never happens. Furthermore, in the operator log, some of events show `%!s(MISSING)` as the node name.

      Attachments

        For Gerrit Dashboard: K8S-542
        # Subject Branch Project Status CR V
        98531,2 K8S-542: Only skip recovery of a down node if it's pod is over 60 seconds old master couchbase-operator Status: MERGED +2 +1

        Activity

          People

            mikew Mike Wiederhold [X] (Inactive)
            korrigan.clark Korrigan Clark (Inactive)
            Votes:
            0 Vote for this issue
            Watchers:
            1 Start watching this issue

            Dates

              Created:
              Updated:
              Resolved:

              Gerrit Reviews

                There are no open Gerrit changes

                PagerDuty