Slow server cause vhang!

[Milton Argueta_1]

Someone can help me!
rp5470 O.S.:hp-ux 11i Processors:2, 650MHz and RAM:4GB.
The server is extremly slow and when I execute top command, vhang is the second process consuming resources.
I checked the parameter kernel of the server dbc_max_pct and dbc_min_pct are seted to 4. (Dymanic Bufer Cache Size as Percent of System RAM Size).
What can I do to solve this problem_?
Can you help me please?

[Milton Argueta_1]

I think the server is very slow due to this process vhang, and I would like to know what can I do to solve the problem. Change the Kernel parameters? or.....????

[Michael Steele_2]

A search on 'vhang' comes up negative for the ITRC and the internet, so I can offer you no advice on 'vhang' unless its unique to you're environment.

However I do offer a performance review if you attach the results of the following commands:

sar -u 5 5
sar -d 5 5
sar -v 5 5
sar -b 5 5
vmstat 5 5
swapinfo -tam

[Patrick Wallek]

I think he means 'vhand'.

If 'vhand' is using lots of CPU then your system is most likely paging out (ie. swapping).

Check 'swapinfo -tam' output. If you are indeed swapping then you need to figure out why.

What applications run on this macihne? Is it a large database?

We really need more details.

[Milton Argueta_1]

Here you are the output of the command you recomend me. Exactly the name is "vhand".
and I have a Oracle Data Base running.
O.S.:hp-ux 11i

[Patrick Wallek]

Oh my yes!!!! Your machine is definitely swapping. You have used 100% of 1 of your swap areas and 81% of the other. That is definitely not good. You really need to look at what is using all of your RAM. The first thing I would check is the database SGA size. Then see if there are any scripts, programs, SQL Code that is running that can be (re)written to manage memory better.

With as much swapping as you are doing though, I would highly recommend that you significantly increase the RAM in that machine. Since you are using almost 8GB of swap, I would consider upping the RAM to at least 12GB, if the rp5470 will support. I don't recall the maximum RAM that the rp5470 can have.

[Michael Steele_2]

Well, you've got some HUGE problems here.

This server is severely constrained.

A) Change your swap priorities all to 1 in /etc/fstab:

dev 4096 3328 768 81% 0 - 1 /dev/vg00/lvol2
dev 4000 4000 0 100% 0 - 0 /dev/vg00/swap2

B) Buy more RAM immediately if you can, if you can't add more 2ndary swap. You're out.

total 11286 11050 236 98% - 0 -

C) You have a severe disk or tape bottleneck for you CPU averages 81% wait time.

C2) Any tape work going?

D) Disks c1t2d0 and c2t2d0 are 100% busy. What's on c1t2d0 and c2t2d0? O/S and swap?

D2) Move all of your 2ndary swap off of O/S disks and out of vg00. (* /dev/vg00/swap2 *)

In short, you're O/S is madly swapping out processes so buy some more RAM or turn off some processes.

[Milton Argueta_1]

Thanks a lot for your comments and help.
I have to work very hard inmediatly.
Have a good night! C U later friends!

[Dietmar Konermann]

You are running 11.11... and the kernel is definitely suffering from memory pressure.

Before looking any further, be sure to have these patches installed:

PHKL_28695
PHKL_28410
PHKL_29110

Believe me, installing these on 11.11 may save you a lot of time.

Best regards...
Dietmar.