Performances Issues. Multiples locks.

Home Page › Forums › BizTalk 2004 – BizTalk 2010 › Performances Issues. Multiples locks.

Viewing 1 reply thread
  • Author
    Posts
    • January 22, 2007 at 9:28 AM #17355
      community-content

        Hello,

        I have performance problems with my Biztalk platform.
        Platform is as follow:

        – All BZT DB hosted on a SQL2005 Cluster,
        – 2 App servers in the same BZT groups, not clustered.
        – Most of the Hosts (but FTP ones) have a instance on each server.
        – I get inbound from FTP / File / SOAP, and outbound is to File / SQL / SOAP

        When both server are running, there are a lot of locks contentions (Process 1 locked by process2, p2 locked by p3, –, Pn-1 locked by Pn, Pn locks P1), so dead loop for performances avoiding deadlock mechanism to fire.
        this generate spool to build up at heavy ryhtm

        SQL perf looks Ok (bo obvious disk contention, no CPU 0r memory crazy usages.. evertyhing ok), and problem disappears when I stop one of the app server.

        Where shall I look for my issue ? Orchestration running within the host ? Not have both servers running all the instance ? any perf counters that might help ?
        I've been digging into doc & google, but not found a fix yet.

         

        Thanks.

        Chris

      • January 23, 2007 at 9:10 PM #17384
        community-content

          Are all your SQL Agent jobs running without any errors?

           

          Are you stressing the system or do you see this under light load?

           

          • January 23, 2007 at 11:57 PM #17385
            community-content

              Everything agents run every minutes without errors. Problem occur under relativly light load (I'm not in a stress-load mode).

              • January 24, 2007 at 9:46 PM #17409
                community-content

                  Chris,

                  This is out of the performance doc. of BTS 2006 from Microsoft, might be of help to you:

                  "Scaling out the message box from one to two SQL Server computers does not improve throughput because in this scenario, one server is dedicated as the primary message box. Every BizTalk Server group must have only one primary message box.

                  The primary message box handles all subscriptions and message routing, while other message boxes handle message publishing. Instance subscriptions (subscriptions that route messages to already-running instances on specific message box servers) are maintained on the primary message box and on the specific message box that hosts the particular instance.

                  To alleviate lock contentions, message publishing is disabled on the primary message box and only the other message boxes can publish messages. Therefore, overall system throughput does not improve because there is still only one server publishing messages.

                  Running multiple message boxes requires distributed transactions. Therefore, overall performance decreases from the additional network-traffic overhead."

                  Regards,

                  Sajid

                   

                  • January 25, 2007 at 11:12 AM #17428
                    community-content

                      Actually, I only have 1 mailbox DB and 2 App server. I did found a possible lead, I have 19000 message in the spool table, and 100 000+ in the suspended queues…. none of them showing up in the BZT admin tools (only 10 suspended instances).

                       I opened a Suppor Request at MS to clean up this up.

                       Chris

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