Recently I ran into some I/O overload problrms on some of our machines and I have been searching to find thresholds for some metrics like:
avgqu-sz await svctm
Then I saw this post on Stack Overflow: display current I/O
I wonder if there is any way to find max queue length of my machine so when
avgqu-sz is more than that number it means overload occurred?
Does queue length depends on RAID?