Info is from here
You can easily see what process is consuming the most I/O bandwidth:
dstat -ta --top-bio dstat -ta --top-io
The first being the top block I/O, and the second the top I/O (including cached). This may help you validate if this is what you expect (both as consumer as well as the consumption).
There are some other plugins that can help you identify the disks utilization rate:
dstat -tdf --disk-util
So if you have I/O bottlenecks, this will tell you which disk(s) are involved so you can see if something can be done about it at that end.
http://pkgs.repoforge.org/dstat/?C=M;O=D
http://dag.wieers.com/home-made/dstat/