Uploaded image for project: 'Couchbase Server'
  1. Couchbase Server
  2. MB-7279

XDCR error: "database out of sync" seen on source after there's been a cluster change (failover/rebalance) on the destination

    Details

      Description

      c1: 10.1.3.237, 10.3.2.54, 10.3.121.126
      c2: 10.1.3.238, 10.3.2.55, 10.3.121.127

      default: c1 --> c2

      • Load 10K items on c1, items replicate to c2.
      • Failover a node (10.3.2.55) on c2, add back, then rebalance
      • Failover a node (10.3.121.127) on c2, rebalance.
      • Add node 10.3.121.127 to c2, rebalance
      • Add more items on c1.
      • Xdcr error (continuously) seen on c1:
        2012-11-27 17:09:13 - Error replicating vbucket 602: <<"Target database out of sync. Try to increase max_dbs_open at the target's server.">>
      1. 10.1.3.237-8091-diag.txt.gz
        3.48 MB
        Abhinav Dangeti
      2. 10.1.3.238-8091-diag.txt.gz
        7.59 MB
        Abhinav Dangeti
      3. 10.3.121.126-8091-diag.txt.gz
        4.51 MB
        Abhinav Dangeti
      4. 10.3.121.127-8091-diag.txt.gz
        3.21 MB
        Abhinav Dangeti
      5. 10.3.2.54-8091-diag.txt.gz
        5.29 MB
        Abhinav Dangeti
      6. 10.3.2.55-8091-diag.txt.gz
        5.32 MB
        Abhinav Dangeti
      1. Screen Shot 2012-11-27 at 5.12.40 PM.png
        175 kB
      No reviews matched the request. Check your Options in the drop-down menu of this sections header.

        Activity

        Hide
        junyi Junyi Xie (Inactive) added a comment -

        This is expected error raised from XDCR when the target instance startup time does not match the one stored in current vb replicator. In your testcase, it is caused by failover/rebalance/etc. Consequently the checkpoint will fail and current vb replicator will crash, and after 30 seconds by default, a new vb replicator will start to fetch the new target instance start up time.

        This is not a bug.

        Show
        junyi Junyi Xie (Inactive) added a comment - This is expected error raised from XDCR when the target instance startup time does not match the one stored in current vb replicator. In your testcase, it is caused by failover/rebalance/etc. Consequently the checkpoint will fail and current vb replicator will crash, and after 30 seconds by default, a new vb replicator will start to fetch the new target instance start up time. This is not a bug.
        Hide
        junyi Junyi Xie (Inactive) added a comment -

        Not a bug.

        Show
        junyi Junyi Xie (Inactive) added a comment - Not a bug.

          People

          • Assignee:
            junyi Junyi Xie (Inactive)
            Reporter:
            abhinav Abhinav Dangeti
          • Votes:
            0 Vote for this issue
            Watchers:
            0 Start watching this issue

            Dates

            • Created:
              Updated:
              Resolved:

              Gerrit Reviews

              There are no open Gerrit changes