1748169 Members
4307 Online
108758 Solutions
New Discussion юеВ

disk bottle neck

 
SOLVED
Go to solution
skt_skt
Honored Contributor

disk bottle neck

HP-UX 11.00;9000/800/N4000-75; yearly patching done.

disk bottle neck reported followin gis the report from glance(d option)

Req Type Requests % Rate Bytes Cum Req % Cum Rate Cum Byte
--------------------------------------------------------------------------------
Local Logl Rds 1405 94.4 312.2 516gb 11479733 93.4 1087.4 516gb
Logl Wts 84 5.6 18.6 24.4gb 817489 6.6 77.4 24.5gb
Phys Rds 2566 91.7 570.2 254.5mb 8443197 86.3 799.7 526gb
Phys Wts 231 8.3 51.3 11.4mb 1344725 13.7 127.3 25.6gb
User 2761 98.7 613.5 265.7mb 9632966 98.4 912.4 551gb
Virt Mem 2 0.1 0.4 2kb 6236 0.1 0.5 31.2mb
System 34 1.2 7.5 120kb 143350 1.5 13.5 477.8mb
Raw 0 0.0 0.0 0kb 5370 0.1 0.5 5.2mb
Remote Logl Rds 0 0.0 0.0 0kb 0 0.0 0.0 0kb
Logl Wts 0 0.0 0.0 0kb 0 0.0 0.0 0kb
Phys Rds 0 0.0 0.0 0kb 0 0.0 0.0 0kb
Phys Wts 0 0.0 0.0 0kb 0 0.0 0.0 0kb

My obervation is "Cum Byte" column is reporting almost 500Gb for both Logl Rds as well as Phys Rds. Compared to the old stat i collected there is a huge increment.

one of the old status

Req Type Requests % Rate Bytes Cum Req % Cum Rate Cum Byte
--------------------------------------------------------------------------------
Local Logl Rds 4661 89.0 896.3 118gb 2398986 83.6 916.9 118gb
Logl Wts 578 11.0 111.1 9.90gb 469619 16.4 179.4 9.91gb
Phys Rds 4665 83.5 897.1 250.8mb 1811536 73.9 692.4 116gb
Phys Wts 922 16.5 177.3 24.3mb 639904 26.1 244.5 10.3gb
User 5567 99.6 1070.5 275.0mb 2415339 98.5 923.1 126gb
Virt Mem 0 0.0 0.0 0kb 1452 0.1 0.5 2.5mb
System 16 0.3 3.0 64kb 33251 1.4 12.7 107.6mb
Raw 4 0.1 0.7 4kb 1398 0.1 0.5 1.4mb
Remote Logl Rds 0 0.0 0.0 0kb 0 0.0 0.0 0kb
Logl Wts 0 0.0 0.0 0kb 0 0.0 0.0 0kb
Phys Rds 0 0.0 0.0 0kb 0 0.0 0.0 0kb
Phys Wts 0 0.0 0.0 0kb 0 0.0 0.0 0kb

How we can identify a disk bottle neck and which could cause sloe ness to the oracle database processes.

8*750MHHZ cpu/12GB MEM/29GB/swap
4 REPLIES 4
Steven E. Protter
Exalted Contributor
Solution

Re: disk bottle neck

Shalom,

http://www.hpux.ws/system.perf.sh

There is one line in the sar script especially for disk issues. You can run the whole thing or part of it.

It was recently updated by me so it should be considered in a beta status at this time.

It will show you disk that is overloaded. Then you can look at your system layout with lvdisplay or vgdisplay or both and figure out which application is causing the problem.

You should be able to drill in and id the disk with glance, which you are obviously using.

BTW this post was originally found in the Linux section. You'd get about 4 times the eyeball's by posting it to the correct section.

SEP
Steven E Protter
Owner of ISN Corporation
http://isnamerica.com
http://hpuxconsulting.com
Sponsor: http://hpux.ws
Twitter: http://twitter.com/hpuxlinux
Founder http://newdatacloud.com
Ivan Ferreira
Honored Contributor

Re: disk bottle neck

Check if your system is not paging/swapping. Maybe the increase of disk activity is due to memory shortage.
Por que hacerlo dificil si es posible hacerlo facil? - Why do it the hard way, when you can do it the easy way?
skt_skt
Honored Contributor

Re: disk bottle neck

CPU/MEM/SWAP utilization is mostly less than 50% used
skt_skt
Honored Contributor

Re: disk bottle neck

wronng furum