Cisco ACE: Rserver Average CPU core loads via SNMP

[raul.ocana]

hi, 

I have cisco ace4710
I want to monitor rserver windows with the native SNMP service to configure predictor method in base of charge of CPU.

Each of this VM have at least 8 CPU but It seems that the OIDs for processor load don't follow any rule, they appear random. And what's worse they seem to change for no apparent reason and at random times between reboots.

This is the actual setup that I have to probe CPU:

probe snmp CPU
  port 161
  interval 50
  passdetect interval 30
  version 2c
  community XXXX_XXXX_XXX
  oid .1.3.6.1.2.1.25.3.3.1.2.1
    threshold 80
    weight 2000
  oid .1.3.6.1.2.1.25.3.3.1.2.2
    threshold 80
    weight 2000
  oid .1.3.6.1.2.1.25.3.3.1.2.3
    threshold 80
    weight 2000
  oid .1.3.6.1.2.1.25.3.3.1.2.4
    threshold 80
    weight 2000
  oid .1.3.6.1.2.1.25.3.3.1.2.5
    threshold 80
    weight 2000
  oid .1.3.6.1.2.1.25.3.3.1.2.6
    threshold 80
    weight 2000
  oid .1.3.6.1.2.1.25.3.3.1.2.7
    threshold 80
    weight 2000
  oid .1.3.6.1.2.1.25.3.3.1.2.8
    threshold 80
    weight 2000

--
..one of these RSERVER that I have added to SFARM shown this OID for CPU, command snmpwalk:

HOST-RESOURCES-MIB::hrProcessorLoad.2 = INTEGER: 1
HOST-RESOURCES-MIB::hrProcessorLoad.3 = INTEGER: 0
HOST-RESOURCES-MIB::hrProcessorLoad.4 = INTEGER: 0
HOST-RESOURCES-MIB::hrProcessorLoad.5 = INTEGER: 1
HOST-RESOURCES-MIB::hrProcessorLoad.6 = INTEGER: 0
HOST-RESOURCES-MIB::hrProcessorLoad.7 = INTEGER: 0
HOST-RESOURCES-MIB::hrProcessorLoad.8 = INTEGER: 0
HOST-RESOURCES-MIB::hrProcessorLoad.9 = INTEGER: 0

And other RSERVER:

HOST-RESOURCES-MIB::hrProcessorLoad.1 = INTEGER: 1
HOST-RESOURCES-MIB::hrProcessorLoad.2 = INTEGER: 3
HOST-RESOURCES-MIB::hrProcessorLoad.3 = INTEGER: 0
HOST-RESOURCES-MIB::hrProcessorLoad.4 = INTEGER: 0
HOST-RESOURCES-MIB::hrProcessorLoad.5 = INTEGER: 0
HOST-RESOURCES-MIB::hrProcessorLoad.6 = INTEGER: 0
HOST-RESOURCES-MIB::hrProcessorLoad.7 = INTEGER: 0
HOST-RESOURCES-MIB::hrProcessorLoad.8 = INTEGER: 0

Is there a way I could definethis so it doesn't matter what number Windows gives them? 
Any TCL Script to take average CPU core loads via SNMP? any example..?

thanks!

[Kanwaljeet Singh]

Hi Raul,

Are you using predictor least-loaded probe <snmp probe> under serverfarm? Please have a look at the link:

http://www.cisco.com/c/en/us/td/docs/interfaces_modules/services_modules/ace/vA4_1_0/configuration/slb/guide/slbgd/rsfarms.html#wp1023940

Regards,

Kanwal

[raul.ocana]

Hi Kawal, 

Yes, I'm using predictor least-loaded probe MY_PROBE. with the behavior I have attached, if probe cannot read one of this OID of hrProcessorLoad.1,2,3,4,5,..9,  automatically LB put RSERVER out of service because starts OID in  hrProcessorLoad.2,3,4,5,6,7,8,9 out of service.