none
The process is running and waiting for a response from the server

    Question

  • I have tried to reinitialize a replication job a number of times and it keeps failing in the same place on the same table during the bulk upload (141919 rows). The Publication is coming from a SQL 2008 sever to two Subscribers a SQL 2005 and a SQL 2008. The SQL 2005 subscriber was able to synchronies however the SQL 2008 server was not.


    The error I am getting is "The Process is Running and is waiting for a response form the server".

    Based on some recommendations I saw I increased the Agent Profiles Query Timeout to 9000 and LoginTimeout to 90. Still no luck. I have restarted the SQL services on the Subscriber  and reinitialized from a new snapshot, same problem.

    Where do I go from here?

    I checked  MSrepl_Errors and there are no errors from the time I started until the time I had the issue

    I have checked for deadlocks with the deadlock script and there seems to be none.

     

    • Edited by btrowdy19 Wednesday, January 25, 2012 1:06 AM
    Wednesday, January 25, 2012 12:57 AM

Answers

  • I've found this error to be transitory most of the time.

    It usually occurs when bcp'ing a large table during the snapshot process and clears up when it finishes.

    You can run Profiler to see what is going on.  You can also look at Resource Monitor and see if the subscription database is being written to at that time.


    Brandon Williams (blog | twitter)
    • Marked as answer by btrowdy19 Wednesday, January 25, 2012 5:51 AM
    Wednesday, January 25, 2012 2:28 AM

All replies

  • I've found this error to be transitory most of the time.

    It usually occurs when bcp'ing a large table during the snapshot process and clears up when it finishes.

    You can run Profiler to see what is going on.  You can also look at Resource Monitor and see if the subscription database is being written to at that time.


    Brandon Williams (blog | twitter)
    • Marked as answer by btrowdy19 Wednesday, January 25, 2012 5:51 AM
    Wednesday, January 25, 2012 2:28 AM
  • I believe you are right, it has now passed where it got stuck. It took all of 1.5 hours for the SQL 2005 Subscription Machine to replicate all the data and complete the sync. This SQL 2008 server has taken 8 hours already and it's still going. Both subscription servers are in the same data center so latency is not a facor. I might look at disk IO and see if anything is going on this SQL 2008 machine.

     

    Updated: I checked disk performance and it's 65mb/s read 122mb/s write located on a SAN. Plenty of free disk space and memory. Server does not look to be under any load. There is however an outbound publication that is complaining of issues in the Eventlog. It's in a difference instance so I don't think it has any bearing.

     


    • Edited by btrowdy19 Wednesday, January 25, 2012 5:56 AM
    Wednesday, January 25, 2012 5:52 AM