Using IT Analytics to analyze VMware Datastore Performance

This post is the next in my series on using IT Analytics to analyze the performance of our Private Cloud at Bay Dynamics. In the previous post I focused on analyzing CPU and Memory performance for my ESX Cluster. In this post we'll look at Disk utilization metrics for our VMware Datastores.

Again we start with opening our SCOM Performance Cube and filter first to look at only data coming from the nworks VMware MP. That makes it easy to pinpoint the Datastore Performance Rules, including Disk Free (KB) and Disk Percentage Used. Adding in the Minimum, Average, and Maximum measures as well as the entity names for the Datastores, I arrive at the following view:



I loved the CPU and Memory charts from the previous post, so I'll use the same approach here. A few clicks, and we I now have the following view:



Great insight here. By visually analyzing the Disk Space Free and Disk Percentage Used, I can see that Datastores ESX1 and ESX2 have stayed consistent over this time period, whereas ESX3 and ESX4 both saw their free space go essentially to zero at some point during the time period. Its also very clear that the bulk of the remaining disk capacity is on ESX5. I'd like to drill one level deeper though and see how much space each of those Datastores still have available, as well as see if there's any additional insight to be gained by observing the change in space used over time. After adding in the Date attribute and drilling in specifically on the diskFree counter, I can now see:



It looks like we actually freed up some space over the last few weeks. I had been asking my team to keep an eye out for non-production class VMs that we could shift out of my expensive ESX infrastructure over to our less expensive ESXi hosts to make room for upcoming initiatives. I can now see the progress with my own eyes!