A good starting point is to monitor storage, network, and host bus adapter (HBA) performance statistics from the level of the host server. If you're running on Windows Server 2003, you can use Performance Monitor to track, for example, statistics for both the Physical Disk and Logical Disk objects.
Look for overall utilization, as well as the number of I/O operations per second (IOps). Details such as latency and the disk queue length can help you determine whether your current storage design is meeting requirements.
As far as automating the process, you can use enterprise management tools (whether or not they are virtualization-aware) to monitor SAN performance. These solutions may be available from your vendor, and will allow you to isolate the source of performance problems. Potential areas include the host server, the guest OS, the HBA, the storage switches, and the physical disks. I hope this helps to move you along the right path. Good luck!
This was first published in December 2007