Uploaded image for project: 'Couchbase Server'
  1. Couchbase Server
  2. MB-55557

[CDC] History of a dropped collection is made visible to backfill - replica rejects and fails rebalance with an unknown collection error

    XMLWordPrintable

Details

    Description

      Steps To Recreate:

      • Create a 3 node cluster
        Create a magma buckets with (history_retention_collection_default=true,bucket_history_retention_seconds=0,bucket_history_retention_bytes=96636764160, magma_key_tree_data_block_size=131072,magma_seq_tree_data_block_size=9092) (vbuckets = 16, replicas = 2)
        Create fifty collections(total collections including default are fifty one)
        Create 100000 docs in each of the non default collections
        Update all the docs created in above step(16 times)
        Change bucket replicas == 1 and rebalance(rebalance is successfull)
        Again change replicas == 2 and start rebalance
        while rebalance is going on create/delete few collections
        Trigger Compaction
        Stop rebalance and change enable bucket level ttl and set it to 10seconds
        Resume rebalance
        Stop Rebalance and change bucket durability level to majorityandPersistActive
        Resume Rebalance
        Stop rebalance and increase bucket history_retention_bytes to 960GB
        Resume Rebalance
        Stop rebalance and decrease bucket_history_retention_bytes to 2GB(total data on disk at this stage was 110 GB)
        Resume Rebalance
        Rebalance failed with dcp_wait_for_data_move_failed

      Rebalance Failure:

      Rebalance exited with reason {mover_crashed,
      {unexpected_exit,
      {'EXIT',<0.10923.99>,
      {{dcp_wait_for_data_move_failed,"default",3,
      'ns_1@172.23.107.232',
      ['ns_1@172.23.107.97','ns_1@172.23.107.240'],
      {error,no_stats_for_this_vbucket}},
      [{ns_single_vbucket_mover,
      '-wait_dcp_data_move/5-fun-0-',5,
      [{file,"src/ns_single_vbucket_mover.erl"},
      {line,451}]},
      {proc_lib,init_p,3,
      [{file,"proc_lib.erl"},{line,211}]}]}}}}.
      Rebalance Operation Id = 52b233838e709215285e35761577a4f7
       
      Worker <0.10912.99> (for action {move,{3,
      ['ns_1@172.23.107.232',
      'ns_1@172.23.107.97'],
      ['ns_1@172.23.107.232',
      'ns_1@172.23.107.97',
      'ns_1@172.23.107.240'],
      []}}) exited with reason {unexpected_exit,
      {'EXIT',
      <0.10923.99>,
      {{dcp_wait_for_data_move_failed,
      "default",
      3,
      'ns_1@172.23.107.232',
      ['ns_1@172.23.107.97',
      'ns_1@172.23.107.240'],
      {error,
      no_stats_for_this_vbucket}},
      [{ns_single_vbucket_mover,
      '-wait_dcp_data_move/5-fun-0-',
      5,
      [{file,
      "src/ns_single_vbucket_mover.erl"},
      {line,
      451}]},
      {proc_lib,
      init_p,3,
      [{file,
      "proc_lib.erl"},
      {line,
      211}]}]}}}
      

      QE-TEST:

      git fetch https://review.couchbase.org/TAF refs/changes/62/186462/5 && git checkout -b change-186462 FETCH_HEAD
      guides/gradlew --refresh-dependencies testrunner -P jython=/opt/jython/bin/jython -P 'args=-i /tmp/temp_vol_3.ini -p bucket_storage=magma,bucket_ram_quota=256,init_loading=False,bucket_eviction_policy=fullEviction,rerun=False -t storage.magma.magma_crash_recovery.MagmaCrashTests.test_magma_failures_with_CDC,nodes_init=3,skip_cleanup=True,num_items=100000,doc_size=1024,batch_size=1000,sdk_timeout=60,log_level=info,infra_log_level=info,key_size=22,num_collections=51,ops_rate=20000,key_type=SimpleKey,vbuckets=16,replicas=2,doc_ops=update,test_itr=50,bucket_history_retention_seconds=0,bucket_history_retention_bytes=96636764160,standard_buckets=1,magma_buckets=1,num_scopes=1,induce_failures=False,history_for_collections=True,magma_key_tree_data_block_size=131072,magma_seq_tree_data_block_size=9092 -m rest'
      

      Attachments

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

        Activity

          People

            ankush.sharma Ankush Sharma
            ankush.sharma Ankush Sharma
            Votes:
            0 Vote for this issue
            Watchers:
            5 Start watching this issue

            Dates

              Created:
              Updated:
              Resolved:

              Gerrit Reviews

                There are no open Gerrit changes

                PagerDuty