DataPower

DataPower

Join this online group to communicate across IBM product users and experts by sharing advice and best practices with peers and staying up to date regarding product enhancements.

 View Only

  • 1.  SNMP Traps for WDP cpu, ram etc

    Posted Thu December 05, 2019 06:37 AM
      |   view attached

    Hi all

     

    We'd like to monitor our WDPs with snmp traps.

    I saw that there are traps for mem, but none for CPU? 

     

    I want a best practice list to capture all that easy to think of events.

    Any full document with ALL WDP EVENT IDS ?

     

    Please advise guys.



    ------------------------------
    Eli Dotan
    ------------------------------


  • 2.  RE: SNMP Traps for WDP cpu, ram etc

    Posted Fri December 06, 2019 12:25 PM
    Hi Eli, I hope you are doing alright! In my quick research I also couldn't find an SNMP trap related to CPU usage. When I had to implement CPU monitoring for a project, we used the OID from the MIBs that translates to either CPU or Load, don't recall exactly what is there. With such OID, the SNMP server was able to constantly query for such value and then determine whether an alert should be triggered (through pre-defined thresholds) or not.

    Bruno

    ------------------------------
    Bruno Rodrigues Neves
    Integration Specialist
    IBM
    Tampa FL
    ------------------------------

    Original Message


  • 3.  RE: SNMP Traps for WDP cpu, ram etc

    Posted Tue January 07, 2020 04:39 PM
    I second what Bruno suggested... Most of my customers use CPU as a trend metric as opposed to an alert metric.  For memory, when you run out, you're basically done.  For CPU at >90% it could just be a spike, so what I've seen is an OID that takes the CPU measurement and then if the CPU is greater that a period of time, then an alert is generated to have someone take a look at it.

    ------------------------------
    Charlie Sumner
    Orlando FL
    ------------------------------

    Original Message


  • 4.  RE: SNMP Traps for WDP cpu, ram etc

    Posted Tue January 14, 2020 08:35 AM

    Same here, I'm not aware of a trap for CPU.

    You rather poll the CPUUsage status provider.
    Or if you want to get it in logs, you can enable Status Log in Throttle Settings and then subscribe to "system" "debug" to get 0x80400051 as per https://www.ibm.com/support/pages/gathering-websphere-datapower-cpu-and-memory-information - not really what you asked for...



    ------------------------------
    August Ziesemer
    ------------------------------

    Original Message