Forum Discussion

Scott_82377's avatar
Scott_82377
Icon for Nimbostratus rankNimbostratus
May 24, 2007

Big-IP Monitoring

I'm completely new to the world of F5. We just recently started using Big-IP LTM. I set up our monitoring solution which is Solarwinds Orion to monitor CPU and RAM usage on the BIG-IP boxes. Unfortunately Orion is reporting the Big-IP boxes as having the CPU and Ram at 100% utilization. I know this is incorrect. Has anyone seen this before, or now how I can get this to report correctly? Thanks for any help.

3 Replies

  • I would first determine if you are using the correct OUD and hopefully you imported the correct F5 Mibs. Secondly make sure you check with the ask F5 knowledge base to make sure you aren't running into the bug on snmp.

     

     

    Hope that helps if you haven't solved it already.

     

     

    /cb

     

  • Hi,

     

     

    You should go to ask F5 and retrieve the document bigip_nsm_guide_943.pdf.pdf which is the BIG-IP® Network and System Management Guide.

     

     

    Then check chapter 16 page 22.

     

     

    You have here a lots of detailled information about how to calculate CPU usage.

     

     

    HTH

     

     

    Extract:

     

     

    You can use SNMP commands with various OIDs to gather and interpret

     

    data on CPU use on the BIG-IP system. Specifically, you can gather and

     

    interpret data for two different graph metrics: TMM CPU Usage and

     

    CPU[0-n].

     

    To gather and interpret the data for this metric, you must perform some

     

    polling and calculations:

     

    • First, you must use the OID to perform two separate polls, at an interval

     

    of your choice.

     

    • Next, you calculate the delta of the two poll values.

     

    • Finally, you perform a calculation on the OID delta.

     

     

    Table 16.13 shows the individual OIDS that you must poll, retrieving two

     

    separate poll values for each OID.

     

     

    CPU[0-n]

     

    sysMultiHostCpuUser (.1.3.6.1.4.1.3375.2.1.7.5.2.1.4)

     

    sysMultiHostCpuNice (.1.3.6.1.4.1.3375.2.1.7.5.2.1.5)

     

    sysMultiHostCpuIdle (.1.3.6.1.4.1.3375.2.1.7.5.2.1.6)

     

    sysMultiHostCpuSystem (.1.3.6.1.4.1.3375.2.1.7.5.2.1.7)

     

    sysMultiHostCpuIrq (.1.3.6.1.4.1.3375.2.1.7.5.2.1.8)

     

    sysMultiHostCpuSoftirq (.1.3.6.1.4.1.3375.2.1.7.5.2.1.9)

     

    sysMultiHostCpuIowait (.1.3.6.1.4.1.3375.2.1.7.5.2.1.10)

     

    TMM

     

    CPU Usage sysStatTmTotalCycles (.1.3.6.1.4.1.3375.2.1.1.2.1.41)

     

    sysStatTmIdleCycles (.1.3.6.1.4.1.3375.2.1.1.2.1.42)

     

    sysStatTmSleepCycles (.1.3.6.1.4.1.3375.2.1.1.2.1.43)
  • I'm having a similar problem with Orion NPM. We are running version 9.0 SP2 and the Big-IP is running version 9.4.4 Build 73.1 Hotfix HF1.

     

    However, Orion NPM is not recognizing the manufacturer/vendor as F5 nor is it able to identify what the machine type is and I am not getting any readings on CPU utilization.

     

    I realize that I need to use some custom pollers with the F5 to get the additional information such as pools and virtual servers but do I also have to use a custom poller to get CPU utilization?

     

    And what do I need to do in order for Orion to determine the manufacturer as F5?