Comparing Performance from Production in test.

  • Hello,

    I am trying to decide what the best course of action would be in regards to testing the performance of SQL in a multi instance cluster in test in comparison to production where the servers are still multi instance but on stand alone servers.

    I would have liked to have log shipped to the test environment believing this would provide a realistic picture as to what the performance would be like if used in the 'real world', but as production and live are isolated from each other I do not think this is possible, and if it was whether this would be the most effective solution. Perhaps someone could confirm this?

    So I was planning on running trace files in production for a period of time, saving them, restoring Live backups to test (that were taken before the traces were started) and then running those on the restored DB's in test. Is this a good way to test? I would be concerned that I had not captured all the activities from live in order to give me an accurate picture.

    Any advice on this would be most valued, or a nod in the right direction.

    Regards,

    D.

  • Duran (10/27/2010)


    Hello,

    I am trying to decide what the best course of action would be in regards to testing the performance of SQL in a multi instance cluster in test in comparison to production where the servers are still multi instance but on stand alone servers.

    I would have liked to have log shipped to the test environment believing this would provide a realistic picture as to what the performance would be like if used in the 'real world', but as production and live are isolated from each other I do not think this is possible, and if it was whether this would be the most effective solution. Perhaps someone could confirm this?

    So I was planning on running trace files in production for a period of time, saving them, restoring Live backups to test (that were taken before the traces were started) and then running those on the restored DB's in test. Is this a good way to test? I would be concerned that I had not captured all the activities from live in order to give me an accurate picture.

    Any advice on this would be most valued, or a nod in the right direction.

    Regards,

    D.

    You cannot exactly clone the real problems from PROD due to network,Hardware, Users volume discrepancies. but YES you can get a close idea on your TEST servers that how any script/SP or whatever will respond in PROD region. there are different tools like load runner , sql profiler , execution paln etc that can help you to troubleshoot the problems.

    -------Bhuvnesh----------
    I work only to learn Sql Server...though my company pays me for getting their stuff done;-)

  • Hello,

    Thank you for taking the time to get back. There are no problems really, I just need to know if the server will take a normal day to day workload as the production boxes do. Can you recommend any particular profiler template or items to add to one?

    Regards,

    D.

Viewing 3 posts - 1 through 3 (of 3 total)

You must be logged in to reply to this topic. Login to reply