DEVICE IDENTIFICATION

[rajesh73]

HOW TO INDENTIFY THE DEVICE utilization

$sar -d 1 1

12:37:14 PM       DEV       tps  rd_sec/s  wr_sec/s  avgrq-sz  avgqu-sz     await     svctm     %util
12:37:15 PM cciss/c0d0      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00
12:37:15 PM cciss/c0d1     38.00     52.00     32.00     39.96      2.93      7.34      2.42     96.50
12:37:15 PM cciss/c1d0      8.00      0.00    240.00     30.00      0.11     14.38      4.50      3.60
12:37:15 PM cciss/c1d0p1      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00
12:37:15 PM cciss/c1d0p2      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00
12:37:15 PM cciss/c1d0p3      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00
12:37:15 PM cciss/c1d0p4      0.00      0.00      0.00      0.00      0.00      0.00      0.00      0.00

$ df
Filesystem           1K-blocks      Used Available Use% Mounted on
/dev/cciss/c1d0p1     20315812   9843680   9423492  22% /
/dev/cciss/c1d0p2     30470144   5563044  23334340  52% /opt
/dev/cciss/c1d0p3     48755500   2110412  44128508   55%/var
/dev/cciss/c1d0p4       101086     25293     70574  27% /boot
tmpfs                 12277428         0  12277428   0% /dev/shm
$

from the above output what is c0d0,c0d1 and c1d0 what this for root file system  or raid disk , please explain

[Matti_Kurkela]

Apparently your system has two SmartArray RAID controllers.

/dev/cciss/cXdYpZ = Controller X, (logical) Disk Y, Partition Z. The numbers X, Y and Z all start from zero.

If the "pZ" part is not listed, then it means the whole-disk device for the appropriate logical disk, i.e. c1d0 = the first logical disk on the second SmartArray controller.

Note that all of these refer to logical disks. In other words, each "logical disk" is a RAID set defined to a particular SmartArray RAID controller. If you want to know about physical disks, you'll need a SmartArray configuration tool like hpacucli.

Your filesystems seem to be all on partitions of the first logical disk of the second controller (c1d0). Based on the single set of sar output, it looks like the second controller can perform quite a bit better than the first one: the c0 controller is running at 96.50% utilization as it has been reading 52 sectors and writing 32 (a total of 84 sectors). On the other hand, the c1 controller has read 240 sectors at the same time, and is not even breaking a sweat at 3.60% utilization.

Based on only the information you posted, it is not possible to know what is using the c0 controller and its disks, and why the whole-disk device c1d0 shows significant utilization although the per-partition statistics are all zeroes.

Are you running a database that could be using logical disks in raw mode? Or is there some kind of a disk-based backup or disk-cloning operation running? Any of these might be accessing the disks through the whole-disk device, causing the per-partition statistics to remain at zeroes.