Uploaded image for project: 'Couchbase Server'
  1. Couchbase Server
  2. MB-36794

Ephemeral: Failover node followed by rebalance is stuck forever.

    XMLWordPrintable

Details

    Description

      1. Create a 4 node cluster (on CentOS Linux release 7.6.1810 (Core))
      2. Create a Ephemeral bucket(replicaNumber=1) and load 250k items with durability=majority
      3. (Hard) Failover 172.23.105.225 and rebalance it out.
      4. During step 3, load another 250k items

      Rebalance is stuck forever at 89%

      Attachments

        3. AfterFailover.png
          1.01 MB
          Ritesh Agarwal
        4. BeforeFailover.png
          988 kB
          Ritesh Agarwal
        8. gdb.txt.220
          138 kB
          Richard deMellow
        9. gdb.txt.221
          138 kB
          Richard deMellow
        10. gdb.txt.223
          54 kB
          Richard deMellow
        11. gdb.txt.225
          21 kB
          Richard deMellow
        For Gerrit Dashboard: MB-36794
        # Subject Branch Project Status CR V
        117804,5 MB-36794 Ensure we disconnect when we receive EINVAL mad-hatter kv_engine Status: MERGED +2 +1
        117876,1 MB-36794 Ensure we disconnect when we receive EINVAL master kv_engine Status: ABANDONED 0 0
        117890,3 MB-36794: Update highest deduped seqno on abort mad-hatter kv_engine Status: MERGED +2 +1
        117951,2 Merge remote-tracking branch 'couchbase/mad-hatter' master kv_engine Status: MERGED +2 +1
        121189,14 MB-37009: Explicitly handle all DCP error codes from a Consumer mad-hatter kv_engine Status: MERGED +2 +1

        Activity

          People

            ritesh.agarwal Ritesh Agarwal
            ritesh.agarwal Ritesh Agarwal
            Votes:
            0 Vote for this issue
            Watchers:
            6 Start watching this issue

            Dates

              Created:
              Updated:
              Resolved:

              Gerrit Reviews

                There are no open Gerrit changes

                PagerDuty