I have noticed when I check out the /proc filesystem, there are a few parameters that show up with respect to disk activity. Unfortinately, they seem to be vmware-specific and there is nothing in the documentation about them. I have some thoughts, but I'd love it if someone in the know could help me out:
in /proc/vmware/scsi/vmhbaX :
paeCmds paeCopies splitCmds splitCopies
I have noticed when people post outputs when they are having storage issues normally have a high number of bus resets AND high values in these fields.
When I encounter a host without issues, these numbers are generally zero.