ISE 2.0 CPU alarm

[alberx]

Hi,

I recently activated the alarms in ISE 2.0 and seems to work fine. But I´m receiving an "High Load Average" alarm I cannot undertand. The description says:

Description :

The ISE system is experiencing high load average.The load average number represents the average system load over a 5-minute period.A load average of 8.00 is 100% CPU utilization on 8-core box. This alarm will be triggered when the load average is above the number of processors available on your system

In the details the load average reporting is 15:

Details :

High Load Average : Server=XVPISExxxxxx : Load Average=15

This could be possible if the CPUs were 2 core, but they aren´t. ISE is a virtual machine assigned with 8 core.

                                XVPISExxxxxxx/admin# sho inventory

                                 NAME: "ISE-VM-K9          chassis", DESCR: "ISE-VM-K9          chassis"

                                PID: ISE-VM-K9         , VID: V01 , SN: OJNEHBMAS2L

                                Total RAM Memory: 32880304 kB

                                CPU Core Count: 8

                                CPU 0: Model Info: AMD Opteron(TM) Processor 6276

                                CPU 1: Model Info: AMD Opteron(TM) Processor 6276

                                CPU 2: Model Info: AMD Opteron(TM) Processor 6276

                                CPU 3: Model Info: AMD Opteron(TM) Processor 6276

                                CPU 4: Model Info: AMD Opteron(TM) Processor 6276

                                CPU 5: Model Info: AMD Opteron(TM) Processor 6276

                                CPU 6: Model Info: AMD Opteron(TM) Processor 6276

                                CPU 7: Model Info: AMD Opteron(TM) Processor 6276

If I get a report of the server it never shows a higher level of cpu than 25% (attached file).

Does anyone knows if this alarm is correctly created in the system or how to create a real one??

[alberx]

I oppened a TAC case and this is the response:

I think we could be matching some defects. To try a workaround for them please:

* Go to Work Centers->Trust Sec-> overview->dashboard on GUI and in Trustsec dashboard please modify the dashboard refresh interval as never.

* After performing this change we would recommend to reload the ISE MNT nodes to remove the old long running cached sessions. After this please monitor this ISE node for high load average alarms.

I will check for some days.

Finally TAC gave me a couple of articles that talk about "load average" values, and why the alarm can report higher values:

http://blog.scoutapp.com/articles/2009/07/31/understanding-load-averages

http://superuser.com/questions/23498/what-does-load-average-mean-in-unix-linux

[andres.picos]

Hi Alberx, The procedure given by TAC finally results? I have this problem too. I have ISE 2.0.0.306.

[Damien Madison]

Was this issue ever resolved?

I'm having the same issue on ver 2.1.0.474

[RandomAnonymousName]

This is also happening on ISE 2.1.