Azure MP part 2: Configuring monitoring
This is part two of our journey towards monitoring Azure resources with SCOM, using the Azure Management pack. In Part 1: Installation, we finished installing the Management Pack and managed to connect one of our subscriptions where the resources we want to monitor are.
Now, let’s move on to configuring the monitoring for actual resources or services.
Step-by-step guide to configuring monitoring for the Azure MP
You should now have an option in the Authoring pane under “Management Pack templates” to add some monitoring. Select the Microsoft Azure Monitoring option and hit next.
Give it a name and save the template in a Management Pack.
Next, you have the option to choose one of the subscriptions you added while installing the MP. Choose the right subscription from the dropdown.
You’ll now notice an option to either monitor resources of a Service Type or resources in a specific Resource Group. Note that choosing Service Type will contain all the resources across all the Resource Groups in that subscription.
I will choose the Service Type mode for this demo.
As you can see selecting the Service Type mode has displayed a warning that the number of service instances found in the subscription is very large (which will be the case for most enterprises). Refer to the MP guide on how do the additional configuration. The guide says:
Increase the number of Maximum CLR threads in the system registry on the Management Server – Done.
Here you will choose the Service Type(s) you want to monitor and the number of instances in that Service Type. Let’s go with the most common one – Virtual Machines.
To add any exclusions, you can do it in the next screen.
I’ll continue without any exclusions, but note that you can also add wildcards in the filter.
Now, this next screen is interesting. Here you get to choose the metrics you want to collect and also set the alerting criteria for them (optional). I will start with collecting a bunch of them and then set alerting for a couple.
You will get a different list of metrics for the type of Service you choose.
Lastly, you get a summary of all the options you’ve chosen throughout the setup wizard.
Click on Create and you’re done!
If you’re monitoring a large number of services, you would recall the prompt we encountered in the above steps. As per the MP guide’s suggestion, we added the registry keys earlier.
Another suggestion I've got is to decrease the frequency of collection to once every 3600 seconds. Now you must make this override for every collection rule that gets created for each metric you collect across every service type. So only collect the ones you really need! (Again – if you feel lost, read the guide!)
Tip – I simply exported the management pack I saved my Azure monitoring template and replaced IntervalSecondsparameter value from 900 to 3600, and imported it back in 😉
Let’s head over to the Monitoring space and see what objects have been populated.
I see in the Monitoring a new folder has been created called “Microsoft Azure” and there are a bunch of views under it. Let’s take a look at some of them!
The Active Alerts view:
Service Performance view:
I see the data for the metrics I chose to collect has started rolling in.
Service State view:
Note on the right that there are some tasks handy to turn the VM on, off or restart it.
Amazing! Almost like a native SCOM object!
I like this. Now let’s try to add another Service type in a different template. I’ll add Storage Account service type this time.
Back to Authoring > Management Pack template > Add monitoring wizard > Microsoft Azure Monitoring.
Uh-oh. Can’t select the same subscription again. This means that to add another Service type, I have to add it in the same template. Let’s go back and edit the one we created for the Virtual Machines Service.
Open the template, and switch over to the Service type tab. Select the other Service type. Storage Accounts in our example.
Now that we’ve added it, time to select the metrics we want to collect.
Switch to the Metrics tab and you’ll see the Storage Account service has been added with its own list of metrics.
Now that we’ve configured the metrics to collect, let’s give it some time to collect them.
Back to Monitoring pane > Microsoft Azure folder. I can see my Storage Accounts have been added in the views.
Storage Account State view:
Storage Account Performance view:
Awesome! You have now successfully configured monitoring for two of your primary Azure resource Service Types, and that wasn’t hard at all!
Now that you have the monitoring configured and all the juicy data flowing into your SCOM management group, all you’re left to do is to show it off with some pretty and useful dashboards for everyone to consume. We'll be covering this in the next part of this series. “But how do you show it off with pretty AND useful dashboards?” So glad you asked. Answer - SquaredUp! 😊