For more information regarding the security incident at F5, the actions we are taking to address it, and our ongoing efforts to protect our customers, click here.

Forum Discussion

spalande's avatar
spalande
Icon for Nacreous rankNacreous
Aug 22, 2013

SNMP OID for tmm cpu utilization

Is there a way to monitor the CPU usage per TMM. OID .1.3.6.1.4.1.3375.2.1.8.2.3.1.28 ,29 ,30 doesn't provide any data.

 

snmpwalk -c public 127.0.0.1 1.3.6.1.4.1.3375.2.1.8.2.3.1.28 F5-BIGIP-SYSTEM-MIB::sysTmmStatTmTotalCycles."0.0" = Counter64: 0 F5-BIGIP-SYSTEM-MIB::sysTmmStatTmTotalCycles."0.1" = Counter64: 0 F5-BIGIP-SYSTEM-MIB::sysTmmStatTmTotalCycles."0.2" = Counter64: 0 F5-BIGIP-SYSTEM-MIB::sysTmmStatTmTotalCycles."0.3" = Counter64: 0

 

snmpwalk -c public 127.0.0.1 1.3.6.1.4.1.3375.2.1.8.2.3.1.29

F5-BIGIP-SYSTEM-MIB::sysTmmStatTmIdleCycles."0.0" = Counter64: 0 F5-BIGIP-SYSTEM-MIB::sysTmmStatTmIdleCycles."0.1" = Counter64: 0 F5-BIGIP-SYSTEM-MIB::sysTmmStatTmIdleCycles."0.2" = Counter64: 0 F5-BIGIP-SYSTEM-MIB::sysTmmStatTmIdleCycles."0.3" = Counter64: 0

 

snmpwalk -c public 127.0.0.1 1.3.6.1.4.1.3375.2.1.8.2.3.1.30 F5-BIGIP-SYSTEM-MIB::sysTmmStatTmSleepCycles."0.0" = Counter64: 0 F5-BIGIP-SYSTEM-MIB::sysTmmStatTmSleepCycles."0.1" = Counter64: 0 F5-BIGIP-SYSTEM-MIB::sysTmmStatTmSleepCycles."0.2" = Counter64: 0 F5-BIGIP-SYSTEM-MIB::sysTmmStatTmSleepCycles."0.3" = Counter64: 0

 

6 Replies

  • uni's avatar
    uni
    Icon for Altocumulus rankAltocumulus

    This is unimplemented on my boxes too.

     

    You would need to walk the processes to find the tmm processes, then use F5-BIGIP-SYSTEM-MIB::sysProcPidStatCpuUsageRecent. to get current CPU

     

  • Hi,

     

    is it a difference to the generally cpu usage?

     

    cpu - usage: the easiest way is to use the Host Mib. if you walk from here: .1.3.6.1.2.1.25.3.3.1, you get the cpu usage for every cpu.

     

    for tmm, I have the same stuff like you, but I dont use it. Its all based on the TMM id. With this id you can combine all the informations you need. PID, CPU index, TMM index, ...

     

    TMM id: .1.3.6.1.4.1.3375.2.1.8.2.3.1.1

     

    So in your example:

     

    TMM 0.0 --> 0% cpu

     

    TMM 0.1 --> 0% cpu

     

    TMM 0.2 --> 0% cpu

     

    TMM 0.3 --> 0% cpu

     

  • is sysTmmStatTmUsageRatio* OID usable?

    [root@B3600-R70-S26:Active:Standalone] config  snmpwalk -v 2c -c public localhost f5 |grep sysTmmStatTmUsageRatio
    F5-BIGIP-SYSTEM-MIB::sysTmmStatTmUsageRatio5s."0.0" = Counter64: 1
    F5-BIGIP-SYSTEM-MIB::sysTmmStatTmUsageRatio5s."0.1" = Counter64: 1
    F5-BIGIP-SYSTEM-MIB::sysTmmStatTmUsageRatio1m."0.0" = Counter64: 1
    F5-BIGIP-SYSTEM-MIB::sysTmmStatTmUsageRatio1m."0.1" = Counter64: 1
    F5-BIGIP-SYSTEM-MIB::sysTmmStatTmUsageRatio5m."0.0" = Counter64: 1
    F5-BIGIP-SYSTEM-MIB::sysTmmStatTmUsageRatio5m."0.1" = Counter64: 1
    

    by the way, you know sysTmmStatTm*Cycles is deprecated, don't you?

    sysTmmStatTmTotalCycles OBJECT-TYPE
            SYNTAX Counter64
            MAX-ACCESS read-only
            STATUS   deprecated
            DESCRIPTION
                    "Deprecated! Please use the provided tmm usage ratios.
                    Traffic management CPU usage. The total cycles spent in traffic management."
            ::= { sysTmmStatEntry 28 }
    
  • oh, nice to know, that it is deprecated.

     

    I don't use it, so I didn't know. For me it is only interesting to know total cpu usage. I don't find a dedicated way to compare to value with a performance value in the gui. Perhaps you can use the statistics view of all your virtual server.

     

    If I look on my device, there is only 1 VS with 1% cpu, all other VS are running with 0% cpu :-) So, I would say sysTmmStatTmUsageRatio* is usable.

     

  • Thanks Nitass for letting me know that sysTmmStatTm*Cycle is deprecated so can you please let me know the formula to calculate TMM CPU using tmm usage ratios. IS there any formula for it?

     

  • can you please let me know the formula to calculate TMM CPU using tmm usage ratios. IS there any formula for it?

     

    as far as i know, it is derived from tm statistics (tmctl output).