Uploaded image for project: 'Couchbase Documentation'
  1. Couchbase Documentation
  2. DOC-24

[Multi-instance testing]CBworkloadgen crashed while running server re-add rebalance during 20Million item insert run.

    XMLWordPrintable

Details

    • Bug
    • Resolution: Fixed
    • Major
    • 3.1.x
    • 2.5.x
    • cli
    • None
    • CentOS 6.4 64-bit
      RAM:256GB
      Dual 2.9GHz 8-core Xeon E5-2690 for 32 total cores (16 + hyperthreading)

    Description

      the following message is seen when an instance is failed over and rebalance is clicked. the instance is not the instance that CBworkloadgen is connected to.
      2014-01-28 11:42:16,919: s1 refreshing sink map: http://localhost:9000
      2014-01-28 11:42:16,927: s0 refreshing sink map: http://localhost:9000
      2014-01-28 11:42:16,927: s2 refreshing sink map: http://localhost:9000
      2014-01-28 11:45:10,348: s2 error: async operation: error: conn.send() exception: [Errno 32] Broken pipe on sink: http://localhost:9000(default@N/A-0)
      2014-01-28 11:45:10,349: s0 error: async operation: error: conn.send() exception: [Errno 32] Broken pipe on sink: http://localhost:9000(default@N/A-2)
      2014-01-28 11:45:10,351: s1 error: async operation: error: conn.send() exception: [Errno 32] Broken pipe on sink: http://localhost:9000(default@N/A-1)
      error: conn.send() exception: [Errno 32] Broken pipe

      Steps to reproduce:
      0) following steps were done to simulate system level tests.
      1) run following command to start inserting 20M items into cluster.
      user1@xxxx-1111 bin]$ ./cbworkloadgen -n localhost:9000 -i 200000000 -t 3 -u Administrator -p password
      2) create groups and hit rebalance.
      3)while rebalance is in progress, hit failover on one node and then add it back. Click rebalance again.
      4)at some point during the above steps, cbworkloadgen crashed.

      2014-01-27 18:52:59,610: s0 warning: received NOT_MY_VBUCKET; perhaps the cluster is/was rebalancing; vbucket_id: 436, key: pymc77480993, spec: http://localhost:9000, host:port: 172.23.100.18:12000
      2014-01-27 18:52:59,610: s0 warning: received NOT_MY_VBUCKET; perhaps the cluster is/was rebalancing; vbucket_id: 439, key: pymc77480108, spec: http://localhost:9000, host:port: 172.23.100.18:12000
      2014-01-27 18:52:59,613: s0 refreshing sink map: http://localhost:9000
      2014-01-27 18:53:48,938: s0 error: recv exception: [Errno 104] Connection reset by peer
      2014-01-27 18:53:48,938: s2 error: recv exception: [Errno 104] Connection reset by peer
      2014-01-27 18:53:48,938: s1 error: recv exception: [Errno 104] Connection reset by peer
      2014-01-27 18:53:48,938: s0 MCSink exception:
      2014-01-27 18:53:48,938: s2 MCSink exception:
      2014-01-27 18:53:48,939: s1 MCSink exception:
      2014-01-27 18:53:48,939: s0 error: async operation: error: MCSink exception: on sink: http://localhost:9000(default@N/A-0)
      2014-01-27 18:53:48,939: s2 error: async operation: error: MCSink exception: on sink: http://localhost:9000(default@N/A-2)
      2014-01-27 18:53:48,940: s1 error: async operation: error: MCSink exception: on sink: http://localhost:9000(default@N/A-1)
      error: MCSink exception:

      Attachments

        No reviews matched the request. Check your Options in the drop-down menu of this sections header.

        Activity

          People

            venu Venu Uppalapati (Inactive)
            venu Venu Uppalapati (Inactive)
            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