CopyQueueLength much too big for your mom

Server | Exchange 2010 RU4 – 4 copies of each DB in the DAG on 4 different servers

What my problem is | I have a CopyQueueLength (CQL) of 2,000,983 for the first replica I created for the DB / My CopyQueueLengh is invalid, too large, huge, not accurate… ETC.. – I am guessing the Copy Queue Length is equal to the highest log number ever committed to the previous replica, and the number is being pulled from clustering service some place. I would like the number to be zero so I can use the replica.

What started my problem | I had a 500GB DB on a 550GB volume

Some problem proof and pretty pictures |

Copy Queue Length in EMC shows as 2,009,168

image

Copy Queue Length in Perfmon show as zero. Not a very exciting graph

image

What I tired and what fixed the issue :

  • Fail the active DB over to the replica – this failed because the CQL was too large
  • Restarted the Exchange replication service on the target machine – No change
  • Restarted the Information store service – no change
  • Deleted the replica and recreating it – no change in the number
  • Suspend the copy, then resume the copy – No change
  • Suspend the copy, then update the copy – No change
  • Restarted the cluster service – Fixed the problem
Related Posts with Thumbnails

About Kevinm