Biztalk queues up all incoming messages

Home Page › Forums › BizTalk 2004 – BizTalk 2010 › Biztalk queues up all incoming messages

Viewing 1 reply thread
  • Author
    Posts
    • #18551

      Hello.

       

       

      We have a couple of Receive locations in two different Biztalk applications running on the same server.  These receive locations are polling different tables in the same database every 3. Second, using stored procedures.  Two orchestrations are subscribed to these receive locations and when a status flag is set in the table,the receive location returns the row to the orchestration and it starts..  The orchestrations call some soap ports,  and perform som more operations that are note very complex.  At the same time, there are other orchestrations in these applications that select data from the tables and update some fields in them. 

       

      At some point in our process, Biztalk seems to stop responding to incoming request, which are all coming through schemas that are exposed as web services.  What happens is the following.

       

      %u00b7         We send a web service request using a web browser

      %u00b7         The web service times out

      %u00b7         In biztalk we see that it collected the message from the web service call but it is only listed as active in the Biztalk Administrator console. No errors what so ever, it just hangs there in the active state.  There are no errors in the event log except timeout errors.

      %u00b7         If we call other webservices that communicate with the same Biztalk server they all end up as active service instances in Biztalk and the Web call times out.

      %u00b7         So for each message that arrives in biztalk it is queued up as a %u201eRunning service instance“ as active

      %u00b7         To get the Biztalk server to function again we need to restart all host instances and sometimes we need to do it two times.

      %u00b7         After we restart all host instances all of the active service instances start again and perform the operations they were supposed to.  Some end up dehydrated or suspended because of internal time outs.

      Any ideas on what might be causing this?  We have been thinking of database locks but we are not putting the database under that much load. Could it be that polling the database for data at the same time as we are performing some operations on it can put the Biztalk Server into this state?

       

      Any ideas are more than welcome J

       

    • #18772

      Have you tried sp_who2 in T-SQL to check database locks?

      http://www.devx.com/tips/Tip/14208

       

Viewing 1 reply thread
  • The forum ‘BizTalk 2004 – BizTalk 2010’ is closed to new topics and replies.