Aggregate CPU Monitoring

[peter.fields]

Hi all,

 

After reading this article https://www.ca.com/us/services-support/ca-support/ca-support-online/knowledge-base-articles.tec1622806.html it seems that the default behaviour of Spectrum should be to aggregate the CPU usage, however we are receiving alarms for each instance.

 

We have a few devices with 4 and 8 CPU's, they all have the default threshold of 85 set but each time a cpu instance (1/8) goes over the 85% threshold we get an alarm. Is this normal?

 

If so, I have created a watch that calculates the average of the instances:

 

( ( ( ( ( ( ( ( NRM_DeviceCPUUtilization.768 + NRM_DeviceCPUUtilization.769 ) + NRM_DeviceCPUUtilization.770 ) + NRM_DeviceCPUUtilization.771 ) + NRM_DeviceCPUUtilization.772 ) + NRM_DeviceCPUUtilization.773 ) + NRM_DeviceCPUUtilization.774 ) + NRM_DeviceCPUUtilization.775 ) / 8 )

 

However it fails on the devices with 4 CPU, is there any expression to say that if the CPU 772 exists, add it to the equation and finally divide by the amount of counted CPU ?

 

Hope the questions are clear.

[Widjaja Sangtoki]

I guess you are running older Spectrum version ?

We normally don't need to create SpectroWATCH to aggregate CPU utilization. Please check the following.

High CPU alerts in Server 

[peter.fields]

Does 10.1.1 count as an old version?

 

Going back to the article I posted above (https://www.ca.com/us/services-support/ca-support/ca-support-online/knowledge-base-articles.tec1622806.html) it states that by default it monitors the aggregate value and the first instance by default.

 

Each device has 8 instances from 768 through to 775 and this morning I have 3 alarms with the instances 775, 768 and 772. If I read the linked page correctly, this shouldn't happen right?

 

If individual CPU monitoring is a new feature since version 10, how would I go about removing it for these devices?

[Sanyasirao Chillarige]

Hi Peter,

 

As Widjaja given a link in that it explains how to enable aggregate monitoring and disable individual instances.

 

If you are unable to open that link, below are the steps that need to be followed.

 

To change the CPU or Memory threshold alarming from a single instance to

the aggregate value of all instances, copy the following entries from the

$SPECROOT/SS/CsVendor/Cabletron/EventDisp file into the

$SPECROOT/custom/Events/EventDisp file and follow the directions on

commenting and uncommenting the appropriate lines.

# Comment out the following lines if you do not want

# CPU Threshold alarms to occur for individual instances

0x00010f01      R Aprisma.EventPairTimeAttr, 0x00010f02, "0x00010f03 -:-",

0x12bce

0x00010f03 E 50 A 2, 0x00010f03

0x00010f02 E 50 C    0x00010f03

 

# Comment out the following lines if you do not want

# Memory Threshold alarms to occur for individual instances

0x00010f04      R Aprisma.EventPairTimeAttr, 0x00010f05, "0x00010f06 -:-",

0x12bcf

0x00010f06 E 50 A 2, 0x00010f06

0x00010f05 E 50 C    0x00010f06

 

# Uncomment the following lines if you want

# CPU Threshold alarms to occur based on the

# aggregate utilization of all the instances

# 0x00010f07      R Aprisma.EventPairTimeAttr, 0x00010f08, "0x00010f09

-:-", 0x12bce

# 0x00010f09 E 50 A 2, 0x00010f09

# 0x00010f08 E 50 C    0x00010f09

 

# Uncomment the following lines if you want

# Memory Threshold alarms to occur based on the

# aggregate utilization of all the instances

# 0x00010f0a      R Aprisma.EventPairTimeAttr, 0x00010f0b, "0x00010f0c

-:-", 0x12bcf

# 0x00010f0c E 50 A 2, 0x00010f0c

# 0x00010f0b E 50 C    0x00010f0c

 

After making the changes to the $SPECROOT/custom/Events/EventDisp file, you

will need to update the SpectroSERVER Event Configuration by doing the

following:

1. Open a OneClick Console as an ADMIN user
2. Find and select the VNM model
3. Click on the Information tab in the Component Detail panel
4. Find and expand the SpectroSERVER Control folder
5. Click on the Update Event Configuration button

 

 

Hope this might help you.

 

Regards

Sunny

[peter.fields]

Hi Sunny,

 

I made the custom event changes as pointed out in yours and Widjaja's answers, however it doesn't seem to have made any kind of change.

 

I am still getting the threshold alarms from CPU instances.

 

Do you have any other ideas to troubleshoot or resolve the problem?

[Widjaja Sangtoki]

Sorry for the delay but the following should help.

How can CA Spectrum be used to monitor and alarm on individual CPU and memory for devices?