Slow response of the server

[UniRock]

Hi All,

My server 11.23, rx8640 is responding slow when I run commands like ll, su etc.
Database running is Orcale.
What could be the issue?
Attaching the few o/p.

Thanks,
UniRock

[Sunny123_1]

Hi

Check out that your disk has gon faulty or not??


Regards
Sunny

[Michal Kapalka (mikap)]

hi,

check the performance with glance / gpm ( graphical), and check if some LVM erroro is not happen.

mikap

[UniRock]

Hi,

There are a number of disks (around 20-25 with EMC).

How can I use glance to check the performance?

Thanks..

[Sajjad Sahir]

check the system performance

glanceplus
or top, vmstat etc...

thanks and regards

Sajjad Sahir

[Michal Kapalka (mikap)]

hi,

if you are using gpm,

there will be monitored everithing,

disk / cpu / mem / swap and if there some botleneck happen it will show you a red od yellow status.

mikap

[UniRock]

Hi All,

I have attached a performance scipt o/p.

Is there any issue with these disks as avwait>avserv at some places.
c0t6d0
c4t6d0

Thanks..

[Sunny123_1]

Hi

Use ioscan command to check that disk has claimed status or no_hw status

or do

strings /etc/lvmtab |grep c0t0d0

check which vg is using that disk
after that do lvdisplay -v |more

check any stale extents are there or not???

Regards
Sunny

[UniRock]

Hi Sunny,

Disks are CLAIMED and both of them are used in VG00....They do not have stale extents.

Thanks..

[Sunny123_1]

Hi

did you check your swap memory or cpu utilisation

swapinfo -tam to get swap memory


Regards
Sunny

[Mel Burslan]

You need to be a little more specific about what "slow response" means.

Is it hanging after you hit enter and before you see the output, or the output od ll is scrolling the screen in bursts/slowly.

also please post the outputs from

dmesg
sar 5 5

Another good thing to check is to open up a separate window and watch the syslog, live, with

tail -f /var/adm/syslog/syslog.log

as you do commands that respod slowly. See if any errors are getting logged into the syslog.

Performance problems are very site specific and without sitting on a terminal on your system, it is very hard to make even an educated guess.

Hope this helps

[Steven E. Protter]

Shalom,

System could be busy UniRock.

Look for memory leaks.

http://www.hpux.ws/?p=8


Look for performance problems:

http://www.hpux.ws/?p=6

SEP

[UniRock]

Hi All,

By slow response I mean when I enter the command it takes a lot of time to display the output.

Performance script is already attached in my first post.

USED % is ZERO here.

Swapinfo

Mb Mb Mb PCT START/ Mb
TYPE AVAIL USED FREE USED LIMIT RESERVE PRI NAME
dev 8192 0 8192 0% 0 - 1 /dev/vg00/lvol2
dev 20000 0 20000 0% 0 - 1 /dev/vg00/lvswap02
localfs 50882 0 50882 0% 54000 0 3 /swap03/paging
reserve - 32768 -32768
memory 69499 39636 29863 57%
total 148573 72404 76169 49% - 0 -
localfs 50882 0 50882 0% 54000 0 3 /swap03/paging
reserve - 32768 -32768
memory 69499 39636 29863 57%
total 148573 72404 76169 49% - 0 -



Thanks....

[doug hosking]

One thing that jumps out at me is that both ll and su deal with user IDs and groups. If these are slow but a command like "date" runs at normal speed, I'd be interested to see if maybe your pwgrd process is unhappy or if there is some other problem with resolving IDs and groups (such as NIS/LDAP configuration problems or equivalent).

Is there an obvious difference in performance of "ls" vs. "ls -l" ?

Is there an obvious difference in performance of chmod vs. chown on the same file?

What is in your /etc/nsswitch.conf file?