Performance on CPU0 Only

[felyjos]

Hello,

Is it possible to get a performance report for the past three days only on CPU0 for an ESX Server? (not all CPUs combined)

Thanks,
Dom

[Vitaliy S.]

Hi Dom,

Only total CPU Usage can be displayed via reports. Could you please tell me how do use CPU utilization only for 1 core?

Thanks!

[felyjos]

Hello,

Not sure how nWorks MP is capturing this information as from the nWork MP I am receiving this alert...

CPU0 on Host servername.domain has exceeded usage threshold by using 99.3333333333333%.

with the following details:

Source: CPU0 CPU0 on Host servername.domain has exceeded usage threshold by using 97.6666666666667%.
Path: servername.domain\CPU\CPU0
Alert Monitor: nworks VMware: ESX Host CPU has exceeded usage threshold

Summary
This CPU instance has exceeded threshold for CPU usage percent

Causes
VMHostCPU-cpuUsedPct: Percentage of time during which the host CPU was busy during the last interval.

Resolutions
Check available CPU resources and configuration (CPU Shares, Reservations etc.). Check how many VMs are running on this Host. Check VM processor affinity. View graphed historical data here

External Knowledge Sources
Refer to VMware's online documentation and the nworks Online Dictionary for more information.

with the following configuration
- <Configuration>
<ObjectName>VMHost-Cpu</ObjectName>
<CounterName>cpuUsedPct</CounterName>
<InstanceName>$Data/Property[@Name='cpuName']$</InstanceName>
<Value>$Data/Property[@Name='cpuUsedPct']$</Value>
<Threshold1>85</Threshold1>
<Threshold2>90</Threshold2>
<EntityIdTarget>$Target/Property[Type="nworks.VMware.VEM.VMHOSTCPU"]/id$</EntityIdTarget>
<ClassName>HOSTCPUSTATS</ClassName>
<NumSamples>1</NumSamples>
</Configuration>

From this knowledge it refers to VMware documentation
So now I need to provide information on a weekly basis about the usage for CPU0 and/or other CPUs...?

Thanks,
Dom

[Vitaly.Mzokov]

Hi Dom,

As Vitaliy has already mentioned, only total CPU Usage can be monitored and displayed on Veeam ONE v6's reports.

And YES, you're correct - By default, nWorks Collector gathers even lower level metrics (e.g. per-CPU Core performance) from its monitored VMware objects.

As for Veeam ONE v6, could you please let me know of the usage example when you do require per-CPU Core monitoring and reporting, so we could understand the main idea behind the scene?

Thanks,
VM.

[felyjos]

Hello Vitaly,

I am getting alerts through nWorks that the CPU0 is overused. So my managment is asking me to get the same information from ate least two other sources to be sure it is not a SCOM/nWorks issue but reallly a VM issue.

Thanks,
Dom

[Vitaliy S.]

Dom, you may want to check CPU0 utilization via vSphere Client to verify that this is not an nworks MP issue. Thanks!