|
||||||||||||||||||||||||||||||||||||||
|
Using
the UNIX iostat utility to find disk bottlenecks
Every Oracle
administrator knows that disk I/O is the number one cause of performance
problems. However, some Oracle folks do not know how to identify disk
bottlenecks. One little-used
UNIX utility can be very valuable for identifying disk I/O bottlenecks.
If you are not using a disk storage array such as EMC, the iostat
utility can show you the actual I/O activity against each of your physical
disks. Here is a
sample of the iostat command: prompt>
iostat 15 The “iostat
15” says to display one line of output for every disk that we have
attached to our database server. In
this case we get output every 15 seconds, each showing the amount of I/O for
each disk. Here is a sample of the
output: Disks:
% tm_act
Kbps
tps
Kb_read
Kb_wrtn hdisk6
1.0
73.8
4.2
8999945
3314136 hdisk7
0.7
86.4
3.1
13790876
629660 hdisk8
0.0
0.6
0.0
90469
2176 hdisk9
0.8
74.4
3.1
9875281
2532636 hdisk10
0.5
54.6
1.7
9108228
3616 hdisk11
0.1
13.7
0.4
2279793
244 Here
are descriptions for the columns:
The
“% tm_act” shows the percentage of active time for the disk, and this
output, when used in conjunction with a utlestat report (report.txt) can
show you disk bottlenecks. If
you like Oracle tuning, you might enjoy my latest book “Oracle Tuning: The Definitive Reference” by Rampant TechPress. It’s only
$41.95(I don’t think it is right to charge a fortune for books!) and you
can buy it right now at this link: http://www.rampant-books.com/book_2005_1_awr_proactive_tuning.htm
|
|
||||||||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||||||||||