Uploaded image for project: 'JBoss Enterprise Application Platform'
  1. JBoss Enterprise Application Platform
  2. JBEAP-18077

(7.3.z) Scaling up multiple pods in parallel from 0 can lead to 2 coordinators

    XMLWordPrintable

Details

    • Bug
    • Resolution: Unresolved
    • Critical
    • None
    • 7.3.0.CD18
    • Clustering, OpenShift
    • None

    Description

      Use case here is set up upstream DNS_PING cluster of 3 nodes and run multiple clients on it. Check that session is properly replicated and counter is correct over whole cluster.

      But it happened counter was changed.

      Attaching pods logs, events and test logs.

      Suspicious to me seems this message in pods - indicating split brain?

      INFO  [org.infinispan.CLUSTER] (thread-6,null,hejbc-1-d7jxc) ISPN000093: Received new, MERGED cluster view for channel ee: MergeView::[hejbc-1-fds24|2] (3) [hejbc-1-fds24, hejbc-1-d7jxc, hejbc-1-9trs5], 2 subgroups: [hejbc-1-d7jxc|1] (2) [hejbc-1-d7jxc, hejbc-1-9trs5], [hejbc-1-fds24|0] (1) [hejbc-1-fds24]

      This can be followup of JBEAP-17083, this particular test was ignored because of JBEAP-17083 until CD18.

      Or this can be related to other Upstream DNS_PING failures we are investigating currently JBEAP-18066 JBEAP-18070

      Attachments

        Issue Links

          is related to

          Feature Request - Feature requests from customers and/or users JBEAP-18443 [Doc] Document how to start EAP cluster safely

          • Critical - To be worked on immediately following BLOCKER issues.
          • Closed

          Bug - A problem that impairs or prevents the functions of the product. JBEAP-18066 DNS_PING rolling update: Client gets "HTTP 500 error: Internal Server Error"

          • Major - To be worked after higher priority work items are addressed. When the severity is high and the effort to change it is low to moderate, this term is chosen. Issues in this category likely have an existing workaround but implementation or execution may be non-trivial.
          • New

          Bug - A problem that impairs or prevents the functions of the product. JBEAP-18070 Upstream DNS_PING lost member during scale up

          • Major - To be worked after higher priority work items are addressed. When the severity is high and the effort to change it is low to moderate, this term is chosen. Issues in this category likely have an existing workaround but implementation or execution may be non-trivial.
          • Closed
          Mentioned in

          [Trello] Card EAP QE Microteam leads: OpenShift TS - cluster tests shouldn't start servers in parallel

          Activity

            People

              rhn-engineering-rhusar Radoslav Husar
              mchoma@redhat.com Martin Choma
              Votes:
              0 Vote for this issue
              Watchers:
              8 Start watching this issue

              Dates

                Created:
                Updated: