Logical Disk monitoring - Overrides galore?

Question

Hi all!
We have a number of systems with differing needs for storage thresholds.  The defaults work the majority of time, but there are a number of 'buckets' our disks need.  For example, some disks might warrant an error at 5GB free, 10GB free, or even
25GB free.  Let's pretend the percentages aren't important for now.
Just to be clear, is it correct that I would need to create 2 (system / non-system) * 3 (2003, 2008, 2012) = 6 overrides for every single bucket I want (and make no mistake, we need way more than 3)?
This seems absurd.  I can handle dumping the appropriate logical disks into buckets based on PowerShell or some other programmatic method.  Creating
overrides seems a bit more complicated.  Is there not a simpler way to do this?
I'm comfortable using PowerShell, but the 600 line example above seems like a pretty steep requirement for a very basic and common task (that is cumbersome and delay ridden when using the SCOM console).
Please tell me there is an easier way to create overrides for logical disk space monitoring!

Answer 1

I run into the same questions and problems with most customers. What I always suggest is to give the power to the server and/or application owners. This can be accomplished a few different ways, and I have done this by using a registry entry or by using
a file on the root of each disk.
For example, everyone gets default thresholds for disk monitoring. If the server owner wants different thresholds for a disk on their server(s), then they would create a csv file on the root of each disk in which they want different thresholds (or create
a registry entry with similar threshold configuration).
If you modify the built-in disk monitor by checking whether this particular file is present, then read the file to get the custom thresholds and bypass the default thresholds for monitoring that particular disk. The built-in disk monitoring script is pretty
long, but if you know a little VBscript then you should be able to figure out where to add more logic to retrieve custom thresholds based on csv file or registry information.
You will need to put this monitor in a new pack (for example, Windows 2008 Operating System Extended), and disable the unit monitor in the vendor pack. This would end up being 2 new "extended" packs if you were to also do this for Windows Server
2012.
This will effectively remove the responsibility from the SCOM admin to manage hundreds (or more) disk monitoring thresholds, and place the responsibility into the hands of the server owner. This has worked quite well in any environment I've work in.
Jonathan Almquist | SCOMskills, LLC (http://scomskills.com)