2 Replies Latest reply on May 21, 2012 4:09 AM by stuartwhyte

    Latency report assistance

    stuartwhyte

      Hi,

       

      Please could anyone help with the SQL for a report that will give the results of the following:

       

      Any node that has an average latency the exceeds [custom property - latency Threashold] for at least 30 minutes in the last 24 hours.

       

      ie. Latency threashold = 10

       

      Time - latency

      00.00 - 5

      00.15 - 5

      00.30 - 10

      00.45 - 5

      01.00 - 10

      01.15 - 10

      01.30 - 10

      01.45 - 5

      02.00 - 5

      02.15 - 5

      02.30 - 5

      02.45 - 5

      02.00 - 5

       

      Would appear in report, whereas

       

      Time - latency

      00.00 - 5

      00.15 - 5

      00.30 - 10

      00.45 - 5

      01.00 - 5

      01.15 - 5

      01.30 - 10

      01.45 - 5

      02.00 - 5

      02.15 - 5

      02.30 - 5

      02.45 - 5

      02.00 - 5

       

      would NOT

       

      This report will be sent as a daily scehdueled task.

       

      If an advanced alert can be generated in the same way that would be a bonus.

       

      thanks in advance

       

      Stuart

        • Re: Latency report assistance
          erica.gill

          Hi Stuart,

           

          Thinking about this you could get a standard report to do 99% of the work for you here. Just create a standard availability report, have it list the response times, etc you need. Then add a filter under the "Filter Results" tab to filter on response times greater than 73. Then if you view the SQL of the report, the 73 should stand out and you can look at replacing the 73 with your SQL for the custom property.

          I haven't tested it myself but in theory it should work and the process is handy for building complex SQL queries.

           

          All the best,

          Erica

          1 of 1 people found this helpful
            • Re: Latency report assistance
              stuartwhyte

              Hi Erica,

               

              Thanks for the pointer on this.  It is a good method, but I am unable to see a way of selecting the two time frames that I require.  I need to see in the report any node that has a 30 minute average greater that [Latency_threshold] at any point in the last 24 hours.  From what I can see, I can only select either 24 hour avergae, or 30 minute average, neither of which produce the desired results.

               

              Thanks

              Stuart