[geeks] A Linux Tech Support Question...

[Jonathan Katz]

Folks,

My google-fu is week and I haven't found an answer. I'm also less of a
sysadmin and more of an application admin these days, too.

I have a system where we have multipathing to a SAN. There are two devices,
/dev/sdb and /dev/sdc which are my LUN (presented twice) and we have a dm-0
device which is the "meta" device for those two LUNs.

Running "iostat -xd 3" to get a grip on what's going on I find that almost
consistently, dm-0 is at (or near) 100% with the load "balanced" among
/dev/sdb and /dev/sdc split 50/50, 70/30, etc. See below (I took out some
extra stuff):

Device:         rrqm/s   wrqm/s   r/s   w/s   rsec/s   wsec/s avgrq-sz
avgqu-sz   await  svctm  %util
sda6              0.00  1254.33 39.33 73.00   605.33 10618.67    99.92
0.15    1.38   1.17  13.17
sdb             141.33   289.67 243.67 29.33 53261.33  2658.67   204.84
4.26   15.73   2.48  67.70
sdc              81.67   422.00 109.33 53.67 25730.67  3805.33   181.20
2.28   14.00   2.24  36.57
dm-0              0.00     0.00 577.00 793.33 79077.33  6346.67    62.34
21.32   15.95   0.73  99.90

The question is how do I get "real" utilization #s from dm-0? Is it always
at/near 100% utilization or is that just a function of multipathing? Is there
a different utility I should be using?

My system is 30-40% "iobound" (iowait as expressed in both vmstat and iostat)
so this 100% figure that I'm seeing could be legitimate. I just wanted to make
sure I'm reading things right before pressing forward with my customer.

Thanks!

-Jon