Hi - You cant isolate the CPU perf by volume name. You can look at the CPU perf using sysstat -m X (X being the number of seconds), eg: sysstat -m 2. All that will show you is the CPU utilization by core.
The iscsi perf looks ok in your screenshot, with read and writes in the thousands, but nothing too crazy. Your Disk Reads are not ! the disk reads are massive at 70/80 kb/s
Your % cache hit is a little low at 90% as well.
What are you running attached to the LUN's/system as its performing a lot of disk reads from your LUN's and this is probably causing the CPU to go through the roof, especially if its a random workload. Your CIFS and NFS Connections are also performing a steady amount of i/o. Could something be streaming to/from a connection?
Thank you, your comment enrich my little knowledge about performance. I figured out what was causing the high CPU processing. It was the deduplication of a volume backup database that began at dawn and went during the day causing a high throughput in storage. Thanks!