How to Monitor Memory to in Percent with two threshold Warning and Critical in SCOM

Question

Hello Folks,

I'm trying to figure out the Performance Counter to configure Custom Monitor for Memory Utilization in % in SCOM 2022. Can someone pls help me to get it.

Regards

Sai

Answer 1

Hi Sai Kumar!

Setting up a custom monitor for memory utilization in SCOM 2022 with warning and critical thresholds involves a few steps.

Here's a guide to help you get started:

1. Create a New Management Pack:
   • Open the SCOM console.
   • Go to the Authoring pane.
   • Right-click Management Packs and select Create Management Pack.
2. Create a New Monitor:
   • In the Authoring pane, expand Management Pack Objects and select Monitors.
   • Right-click and choose Create a Monitor > Unit Monitor.
   • Select Windows Performance Counter > Static Threshold > Double Threshold and click Next.
3. Configure the Monitor:
   • Choose the management pack you created earlier.
   • Name your monitor (e.g., "Memory Utilization Monitor").
   • Select the target (e.g., Windows Server Operating System).
4. Set Performance Counter:
   • For the performance counter, use Memory\% Committed Bytes In Use or Memory\Available MBytes.
   • Configure the thresholds:
     • Warning Threshold: Set this to your desired percentage (e.g., 75%).
     • Critical Threshold: Set this to a higher percentage (e.g., 85%).
5. Configure Alerts:
   • In the Health tab, set the health states for the thresholds.
   • In the Alerting tab, configure the alert settings to generate alerts when the thresholds are breached.
6. Save and Deploy:
   • Save the monitor and deploy it to the appropriate servers.

If you have any specific questions or run into issues, feel free to ask! 😊

Answer 2

Hi,

Regarding the error you're encountering, the issue arises because the low threshold value should indeed be less than the high threshold value. For Memory\Available Mbytes, you should set the low threshold to a smaller value (e.g., 500 MB) and the high threshold to a larger value (e.g., 1 GB). This way, the monitor can correctly trigger alerts when the available memory falls below the specified thresholds.

Answer 3

If your goal is to:

Have an alert when the Higher threshold (1000MB or 1GB) is passed, then this health state would be set for Critical.

Have an alert when the lower value falls below 500MB, this Health State would set to Warning.

If the value stays between 500MB and 1000MB, the Health State would be Healthy.

When this type of setup is used, you should be very familiar with the performance history for the resource(s) this monitor will be used for. If not, the system could see many alerts.

Typically, a Warning and Critical Threshold are set via separate overrides for a certain server or group of servers. So as long as the value does not hit the Warning Health State or Critical Health State, it will be Healthy.

Hope this helps.

Tony

Answer 4

I was able to achieve this with below setup. Low Value-500 and High Value-1000 and healthy state shown in the snap.

This generated alerts when Memory goes below 1GB for Warning and 500 for Critical.

Also yes I have created this monitor for a group of servers.

I had a thought about combination of no of intervals and no of samples hope that option isn't available with Double Threshold. This could avoid temporary spikes.

Also I'm trying to figure out to mention the Available Memory in description. If you could guide that will be helpful.

Thank you very much Tony for all your support.

Answer 5

Morning, the simplest way if you are using the standard SCOM alert formatting would be to:

Create a new Notifications Channel and in the Format Section in the Description Field replace the

Alert Name or Description with what ever you want it to be

$Data/Context/DataItem/ResolutionStateName$ - Alert Name or Description : $Data[Default='Not Present']/Context/DataItem/AlertName$

OR

There should be a Description field in the monitor you created.

Tony