The Evolution of The Top 20 vCenter Metrics

If you haven't noticed this already, the more you learn about VMware, the more you realize how much you don't know. At least that is how things have worked out for me. About two years ago, I created a technical presentation called "The Top 20 vCenter Metrics You Should Care About", which has brought insight to 1000's of VMware users. Little did I realize at the time this would be the beginning of a seemingly never-ending journey to fully understand the key performance metrics in vCenter.

Upon delivering the most recent iteration of this presentation for the New England VMUG last week, I looked upon some familiar faces from years past and began reflecting upon what new treats they would see and what led to these discoveries. Here are some of the thoughts that crossed my mind:

  • Many of the metrics have interesting "quirks" that can throw you for a loop if you're not on your toes. The more lab work I do, the more I find these quirks and add them to my presentation.
  • Measuring memory usage in vCenter is still far from being a perfect science and I'm always striving to make sure I give admins and engineers the most educated approach to this area of performance.
  • I continue to discover which metrics do not have vCenter alarms available for them and which alarms are available but lack the configuration options necessary to actually make them useful. As I find these flaws, I find it important to point them out so admins can try to find a means to fill those gaps.
  • Lastly, I can't express enough appreciation for the people who ask questions or otherwise speak up during these presentations. The amount of use cases and fringe cases for the metrics is staggering. I'm always learning new ways to apply them to the successful management of vSphere.

So, while the 20 most important metrics haven't changed – for the most part – our understanding of them and their usage continues to evolve.