Forum Discussion

Mark_Warner_337's avatar
Mark_Warner_337
Icon for Nimbostratus rankNimbostratus
Oct 18, 2017

Recommended BIG-IP CPU Utilization monitoring thresholds

I am writing a monitoring configuration guide for our BIG-IP Load Balancers. We are running OpManager and currently show 99% CPU utilization on both kinds of balancers (BIG-IP 540 and 3600) with one reading at 96%. With the default thresholds in OpManager these show as being in Critical state, but performance seems normal. What would you recommend for alert/alarm thresholds for CPU Utilization via SNMP?

 

I'm a tech writer, not a network engineer, but my research seems to indicate that CPU Utilization with these devices is a compound value that may not have value as a performance/trouble indicator.

 

application delivery
BIG-IP
devops

4 Replies

Sort By
  • bbukane_324032's avatar
    bbukane_324032
    Icon for Cirrus rankCirrus
    Oct 18, 2017

    Recommended alert/alarm thresholds we use in Network devies is 80-85% in which if device reaches it threshold will receive a alert/mail so that we can check in further in device, what is causing in device/s to go high. Hope I answered your query....

     

    • Mark_Warner_337's avatar
      Mark_Warner_337
      Icon for Nimbostratus rankNimbostratus
      Oct 18, 2017

      Thank you. I'd like to collect a couple of other responses before making a decision, but that seems reasonable.

       

  • BB16's avatar
    BB16
    Icon for Nimbostratus rankNimbostratus
    Oct 18, 2017

    Recommended alert/alarm thresholds we use in Network devies is 80-85% in which if device reaches it threshold will receive a alert/mail so that we can check in further in device, what is causing in device/s to go high. Hope I answered your query....

     

    • Mark_Warner_337's avatar
      Mark_Warner_337
      Icon for Nimbostratus rankNimbostratus
      Oct 18, 2017

      Thank you. I'd like to collect a couple of other responses before making a decision, but that seems reasonable.