VAR full (help!)

[Vogra]

Hi All!
the file /var/stm/logs/sys/diaglogd_activity_log is increasing and the var dir is full.
I have the directory /var/adm/crash/crash.4 related with the "problem" last night. All the things toke place after the reboot last night. The send mail is show message that is impossible to create the file because it already exists.
Can I delete the other directories
/var/adm/crash/crash.? ?

What I have to do?

[James R. Ferguson]

Hi Claudio:

If you do not want have your 'crash' files analyzed, then, yes, you can delete them.

See also this thread for other files to remove to gain space in /var:

http://forums.itrc.hp.com/cm/QuestionAnswer/1,11866,0x3feac6af36b7d5118ff10090279cd0f9,00.html

Regards!

...JRF...

[harry d brown jr]

You can delete them if you don't need to send the data to hp for analysis.

[G. Vrijhoeven]

Claudio,

try man savecrash, this should give you some options.

hope this will help,

Gideon

[Vogra]

ok! I did that...
and what about /var/stm/logs/sys/diaglogd_activity_log file ?

Thanx.

[Thierry Poels_1]

Hi,

if you need space in /var in a hurry you can move the crash files to somewhere else, and decide later if they are usefull to you or not.
BTW, as mentionned by James, there are already a number of threads concerning space recovery in /var.

good luck,
Thierry.

[Santosh Nair_1 · Accepted solution]

If you need to get rid of the file, you can. Just remember to stop the diagnostic subsystem before doing so:

/sbin/init.d/diagnostic stop
rm /var/stm/logs/sys/diaglogd_activity_log
/sbin/init.d/diagnostic start

-Santosh

[ajay kumar rastogi]

Hi !
You can add an entry in crontab for removing the undesired logs.The periodicity can be chosen either on a daily basis or on a weekly basis.

[Vogra]

Santosh,
I did it. After I start diagnostic again, the file start to increase. Is it the problem warning? Can my system is in problem? If yes, what I have to do...?

thanx.

[Santosh Nair_1]

If the log file is growing rapidly then that would indicate that there might be a hardware problem. Check the log files in /var/stm/log/os using the STM logtool utility to determine the cause of the logs. Hope this helps.

-Santosh

[Marcin Wicinski]

Hi,

1. put crash.4 directory on a DDS tape and send it to analysis,
2. use cleanup -c (if you have HPUX 11.0 and -c option is not available install PHCO_24347) - it will remove all patches that were superseded at least times.

Later,

[Frederic Sevestre]

Hi,
In adition of the cleanup -c 1, if you want to commit some patches :
swmodify -x patch_commit=true -f
where file contains the list of the patches you want to commit :
PHxx_xxxxx.*

[linuxfan]

Hi claudia,


Also check your /var/stm/logs/os directory.

But the best thing to do is to check the disk usage in /var and analyze
du -kx /var |sort -n > /tmp/du.out

-Ramesh

[Deshpande Prashant]

Hi
Is your diagnostic version grater than A25.10 (March 2001)? If not upgrade it.

Thanks.
Prashant Deshpande.