This browser is no longer supported.
Upgrade to Microsoft Edge to take advantage of the latest features, security updates, and technical support.
![](data:image/svg+xml, %3Csvg xmlns='http://www.w3.org/2000/svg' height='64' class='font-weight-bold' style='font: 600 30.11764705882353px "SegoeUI", Arial' width='64'%3E%3Ccircle fill='hsl(310.4, 42%, 40%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3ESR%3C/text%3E%3C/svg%3E)
Hi all,
Can someone help me out with steps to create a simple CPU Utilisation Unit Monitor?
Just want to monitor and alert on a server when its CPU goes above 80%.
Am having trouble understanding some options to use.
Object is most probably "Processor".
Counter, most probably % Processor Time
Should i include all instances of the selected counter ticked or removed in the Performance Counter windows?
Is it necessary to create a Performance Rule for CPU first and then create a Unit Monitor for Alerting or just can do with the Unit monitor directly?
I would start by looking for an existing monitor or rule. You likely will find that the OS MP has what you need. You could just set an override and it can provide a reference for creating a custom replacement if needed. My assumption for including all instances is that it will evaluate each separately. You might also simply choose to use total, which is probably your best choice in this situation.
Hi Andrew,
There are a few issues I see with existing CPU monitor "Total CPU Utilization Percentage":
The fact remains that in every environment you would have Servers running CPU at different Levels.
I have some Reporting Servers running at or Above 90% everyday while there are servers that run stable at 20%.
Hence i need to make sperate CPU Monitors for different Groups of my servers.
Just to simply alert when Servers go above a certain Threshold, regardless of Sampling or Queue length etc.
Using the default just seems to complicate things and does not seem that versatile.
Hi @Saiyad Rahim ,
according to KEvin Holman (se the blog post) you canb ovewrite this behvaiour by setting the “CPU Queue Length” threshold to “-1”.
How does CPU monitoring work in the Windows Server 2016 management pack?
https://kevinholman.com/2017/05/13/how-does-cpu-monitoring-work-in-the-windows-server-2016-management-pack/
As per Kevin:
This will result in the equation ignoring the CPU queue length requirement, and make the monitor consider “% Processor Time” only. If you find this is too noisy, you can use the CPU queue length, but use lower value than the default of 15. Another thing to keep in mind, this is a PowerShell script based monitor, so if you want to run this VERY frequently (the default is every 15 minutes) then consider replacing it with a less impactful native perfmon based monitor.
So, you can ztry this approach or go for the standard CPU monitor, based on counter.
----------
(If the reply was helpful please don't forget to upvote or accept as answer, thank you)
Regards,
Stoyan
Thanks for that article Stoyan.
Once again I am disappointed in how SCOM works.
i can use "-1" for CPU Queue Length but what I understand is that the default will generate an Alert after 30mintes of the threshold being breached 3 times in that 30 min window.
Also there is an issue of how do I tell my Level 1 Support what o do with a Prod Server running high CPU and what to do when a Non Prod Server runs on High CPU.
Again SCOM till to date does not have the ability for the user to Add custom Text to alerts for specific objects.
Hi Stoyan,
Can you advice me if this will work for me:
I have to create separate Instructions in the Alert Description for the below Groups of Servers for Performance Alerts like Logical Disk, CPU and Memory:
Prod-General (Non SQL) Server = these are for the Infrastructure Team
NonProd-General (Non SQL) Servers = these are for the Infrastructure Team
Prod-SQL Servers = these are for Infrastructure and DBA Teams
NonProd-SQL Servers = these are for Infrastructure and DBA Teams
Question:
