/var/stm/log/os full

[singis]

Hello all,
My /var is full. When I tried to check what was causing the problem, I found that they r lot of raw log files created in /var/stm/log/os
named log*.raw they are almost 100's of such files in this directory they were created from last April.
Could any one tell me how this log files are generated. Can I delete them.
Thanks,
Smitha.

[Patrick Wallek]

These are raw log files from HP-UX diagnostics.

They can be deleted, but if you are seeing that many files it could indicate a problem. You really should use the STM tools to look at the logs and see what they contain.

[Robert-Jan Goossens]

Hi,

start xstm
# xstm
File --> Administration --> STM Shutdown

move these files to a save place.

/var/stm/logs/sys/activity_log
/var/stm/data/diaglogd_hold_list
and /var/stm/logs/sys/diaglogd_activity_log

# xstm
File --> Administration --> STM Stratup

Regards
Robert-Jan

[Uday_S_Ankolekar]

These files might have created during some hardware diagnostics. You can delete those files if you don't need it anymore.
/sbin/init.d/diagnostic stop
delete files at : /var/stm/logs/os"
/sbin/init.d/diagnostic start

-USA..

[Devender Khatana]

Hi,
These are raw files creating during stm/cstm. These can be easily moved to some other location. But the size of the files should not be very big really. Is it?

HTH,
Devender

[Hanwant Verma_1]

The file which is keep growing on your system resides in /var logical volume.

Delete the file /var/stm/logs/sys/activity_log

You can stop this file from keep growing by typing the following command:

#/sbin/init.d/diagnostics stop

make sure you have logged in as root.

[Raj D.]

Also try this to find out the biggest file to trim..

cd /var/stm

# ls -lR | sort +4 -5nr | more

Also you can check
/var/adm/syslog/mail.log file , it also grows to big enough size, to reduce /var filesytem,

Cheers,
Raj.

[Michael Bluemm_1]

Hallo Singis,
one reason for these many files can be a hardware problem. Check with ioscan.
(a defective disk produced 23 files every day).You can delete all files but the last two or three without any problem.
Bye, Michael

[singis]

Hello guys,
Thanks a lot.
Actually i tried to check the file sizes in /var all are ok except for /var/stm/log/os and when i open it they are lot of log files in it from april. When I tried to read them I could not since they are raw files. I am new to sys administration so I really didnot know from where this files are created and if I can delete them.
So can any of you clearly tell me why they r created and how I can read the log.
Thanks,
Singis.

[Stefan Stechemesser · Accepted solution]

Hi singis,

the logs are created from the process "diaglogd".
diaglogd is part of the online diagnostics.
Always if a driver detects a hardware error (f.e. scsi driver, sdisk driver, fibre channel driver etc.) an error is logged by diaglogd to /var/stm/logs/os/logX.raw.cur.
If the rawX.log.cur reaches 512k it is switched and a new file with X+1 will continue to be the current logfile.

You can read the files with the Support Tools Manager (cstm, mstm or xstm).
For cstm it works as follows:

cstm
...

cstm>ru logtool
...

Logtool Utility>sr
-- Logtool Utility: Select Raw Log File --

Select a raw log file or files to view, format or filter.
[/var/stm/logs/os/log205.raw.cur]
Path: /var/stm/logs/os/
Enter Cancel or File Name:

at this point you can enter any logfile name (full pathname) or press enter to take the current logfile
You will then see a summary of events.

You can see the details with

Logtool Utility>fl
-- Logtool Utility: Format Raw Log File --

Select a directory into which to place the formatted file:
Directory: [/var/stm/logs/os/]

(you MUST choose here a directory in which a formated log will fit (f.e. /tmp). Don't take /var/stm/logs/os as it is full on your server !!!)

normaly, the errors from logtool are also written in clear text in the file(s) /var/opt/resmon/log/event.log

If you have no current hardware problem it is safe to delete or zip outdated logfiles

best regards

Stefan

Stefan

[Fabio Ettore]

Hi singis,

this file can be read by cstm

# cstm
cstm> runutil logtool

They are files strictly of the STM product.
Then you have to discover the reason why they are created searching into log file of STM. You can start the troubleshooting by seeing into syslog.log. Anyway if you tell that they are on last April I think you can follow suggestions and

/sbin/init.d/diagnostic stop
rm /var/stm/logs/os/*
/sbin/init.d/diagnostic start

Also you should check if you have the latest STM version.

HTH.

Best regards,
Fabio

[singis]

Thank you guys. One my disk is corrupted that is the reason I got so many log files.
Thanks for your help,
Singis.

[Hanwant Verma_1]

Can you please assign the points.

Thanks
Hanwant

[Andrew Merritt_2]

>One my disk is corrupted that is the reason I got so many log files.

Yes, that would explain the files. You should have received EMS notifications about the errors, in the /var/opt/resmon/log/event.log file, and via emails to root, if the EMS HW Monitors are running. As well as diaglogd logging to the logX.raw.cur, it notifies the relevant HW monitor (disk_em for disk drives) and the monitor will generate an EMS notification.

Regarding turning off the diagnostics, as suggested by Hanwant above, you wouldn't do this except as a temporary work-around.

And as suggested, make sure that you have up-to-date OnlineDiags installed (though that doesn't sound like a factor in this case).

Andrew