3 Replies Latest reply on Sep 19, 2008 8:48 AM by darryld

    struggling with an advanced alert.



      I'm guessing this is an easy one once I get it right but it's doing my head in now!

       I'm trying to setup an advanced alert like this.

       node name contains 6509


      node name contains 4006


      cpu usgae greater than 75%

      then tick the box for 1 hour so we are alerted if the 6509's or 4000 series chassis we have on the network get stuck with a high cpu for an hour.

      However I try to do this with the alert gui it doesn't seem to work properly, would it be easier to do it as an SQL query but then how do I paste an sql query as an alert trigger?

      Thanks for any pointers,


        • Re: struggling with an advanced alert.
          I think what you need to do is use is a simple condition for the cpu load and then a condition group for the node names. I don't think there is a way to paste a sql query for an alert trigger (at least not in the gui.)  I attached a screen shot of how the trigger would look.  Hope this helps.
            • Re: struggling with an advanced alert.


               Thanks for the reply, thats very similar to what I had tried, the closest I got to what you have suggested was the same as yours but the other way round I had.

               Trigger when any apply

              node name contains 6509

              node name contains 4006

                   Trigger when all apply

                         CPU load is greater than 75

               I've swapped i round to what you have suggested to see if it makes any diference & changed CPU to => instead of just >

               will see what happens - thanks.

               EDIT - thank you - thats working now, I'm not sure what the diference was between your alert and mine because logically they looked the same to - maybe I'm missing something!

                • Re: struggling with an advanced alert.
                  The difference is in the order in which the elements are processed

                  In your original alert the cpu load is assessed first and the 'trigger when all apply' becomes true when cpu is greater than 75%

                  The alert will trigger when either this all group, or either of the node name conditions are true

                  In the working alert the 'trigger when any apply' is assessed first and becomes true when the node name contains either 6509 or 4006 and the alert is triggered when both this and the cpu level is greater than 75% clause is true

                  Try thinking brackets

                  Hope this helps