2 Replies Latest reply on Feb 2, 2018 10:34 AM by cahunt

    Chat with SolarWinds NPM experts now!

    DanielleH

      Follow this link and click on CHAT NOW underneath the Quick Actions menu on the bottom right.  Experts are available now to answer your questions so GO!

       

      20121018_101049.jpg

        • Re: Chat with SolarWinds NPM experts now!
          atripathi

          Dear Team ,

          one of My Client wont to configure alert on memory utilization in three way alert

          1)if memory utilize 30 % then alert will trig for worning

          2)if memory utilization more than 45% then alert will triage for   critical

          3)if memory utilization will revert on below 30 % then alert trig for  recovery

          also give the variaable

           

          host name host ip memory in %

          Balance memory

           

          kindly help to fix this issue on urgent basis

           

          ashish

          Systech

            • Re: Chat with SolarWinds NPM experts now!
              cahunt

              atripathi, you should be able to do this with a Dynamic Query alerting on Status. But first - to set this option into a single alert and solicit node status change you will need to adjust your Thresholds for CPU and Memory on these boxes.

               

              Edit the Node Properties - or select all nodes from the manage nodes page - then edit properties.

              Scroll down and override your Orion thresholds to your 30 and 45. This way you will see the node status change, and can also use the status to alert.

               

              then you can build your alert based on Node Status; and Alert when either Warning or Critical - or even still alert on the actual % usage.

              If you want your Alert to have a status of Warning or Critical itself, then you will need two separate alerts.

               

               

              In your alert either setup to alert based on node status, or RAM/CPU Usage. If you do not separate the conditions either by status or entity then the node status alert will be triggered by either cpu or ram and the alert would need to reset before the other entity would re-trigger the alert. Also, if you get an alert on warning then the change to a critical level or status would not re trigger the alert. But with the thresholds you would see the status change visually. 

               

              You have some consideration regarding how you want the alerts to play out and if you want to alert on both warning and critical status - or just one threshold with the reset under 30.

               

              With either a single or a couple/few alerts on your reset condition make sure you either have status as up, or ramp/cpu use less than 30%.

               

              I would also add some time before trigger and/or reset so you do not get a flapping alert if your ram/cpu use jumps back and forth across that 30% line over a few polls.

               

              Let us know if you need more details, or if you have more questions.

               

              -CharlesH

              Loop1 Systems: SolarWinds Training and Professional Services