Change is the commonest cause of performance and availability problems. We all know that changes planned or unplanned often cause problems even if only for a short time. In shared infrastructure environments this is even more true. Imagine a simple case of adding or removing a VM from an ESX host. Doing either clearly impacts the consumption of resources. So how often like me have you wanted to look at some performance metric and overlay that with alarms or better still configuration changes?
So now I can see what caused the drop off in load In short I can undertake event correlation with performance
So we turned a VM off at this point.