/var filling up - truncating activity_log

[Glenn Mitchell_2]

/var/stm/logs/sys/activity_log is 831MB and growing daily. /var is at 88% with no disk space to expand it to. What is the purpose of this log and how can it be trimmed?
System: HP 9000/800 N4000
OS: HP-UX 11.11

Thank you

[HGN · Accepted solution]

Hi

As the name suggests it is the activity log but to trim the log it is simple

make a copy of this file on some other filesystem and then do
#>activity_log
or #cat /dev/null >activity_log

This will make the file size 0.

Rgds

HGN

[Paul F. Carlson]

This is the activity log for diagmond, the stm daemon. Before clearing out the log file, I would first take a look at it and see why diagmond is logging so much activity. Use cstm or stm to do this.

[Rick Garland]

You will need to use stm to view this file - it is binary format.

[Steven E. Protter]

Let cron fix it.

cp /var/stm/logs/sys/activity_log

> /var/stm/logs/sys/activity_log

Do it as often as needed.

SEP

[Glenn Mitchell_2]

I copied the /var/stm/logs/sys/activity_log to another volume and executed the command #cat /dev/null > /var/stm/logs/sys/activity_log

I am unable to read the System Activity log on either the copied over file or the activity_log back on the /var volume. I get the error message "failed to complete formatting of raw log files...rfer to the ui activity log" and "entry had an unknown tag (1131653896).

The gui display had a yellow indicator on CPU (5d3) 37 "information killed" and a grey indicator on the SCSI disk at 1/12/0/1.9.0. The ioscan of the device shows it is claimed.

When I double click each display, the following details are shown:

-- Current Device Status Log for CPU on path 37 --
The following information was archived during the last hardware scan:
Product : CPU
Qualifier : 5d3
Hardware Path : 37
Device Type : Processor
Device Qualifier : CPU
Most Recent Tool Run : Information
Tool Exit Status : Success
Current Status : Killed
Recommended Action : Check test_activity_log.
Installed tools:
Diagnostic : None
Verifier : None
Exerciser : cpu
Information : cpu
Expert Tool : cpu (Licensed)
Firmware Update : None

-- Current Device Status Log for SCSI Disk on path 1/12/0/1.9.0 --
The following information was archived during the last hardware scan:
Product : SCSI Disk
Qualifier : HP18.2GST318406LC
Hardware Path : 1/12/0/1.9.0
Device Type : Disk
Device Qualifier : Hard
Most Recent Tool Run : None
Tool Exit Status :
Current Status :
Recommended Action : None
Installed tools:
Diagnostic : None
Verifier : disk
Exerciser : disk
Information : scsi_disk
Expert Tool : scsi_disk (Licensed)
Firmware Update : scsi_disk (Licensed)

[Adisuria Wangsadinata_1]

Hi,

You need to restart the STM, here are the steps to reclaim space taken by /var/stm/logs/sys/activity_log file :

Recover the space used by the activity_log file by doing the following:
Stop STM
# xstm
File --> Administration --> STM Shutdown

Move (or remove) the activity_log file
# cd /var/stm/logs/sys; mv activity_log /tmp/activity_log.old

Restart STM
# xstm
File --> Administration --> STM Startup
A new activity_log will now be created. The old file can be removed or archived.

Hope this information can help you.

Cheers,
AW

[Adisuria Wangsadinata_1]

Hi Glenn,

Next time before you close the thread :
assign the point. So everybody end with big smile.

Happy posting.

Cheers,
AW

[Andrew Merritt_2]

This problem is most likely due to an old version of OnlineDiags. There have been some problems in the past that led to this file growing, and they have been fixed.

What version is displayed when you start STM?

To get the latest version of OnlineDiags, go to http://h20293.www2.hp.com/portal/swdepot/displayProductInfo.do?productNumber=B6191AAE

The table of released versions is at http://www.docs.hp.com/en/diag/stm/stm_upd.htm#table.

The latest version for 11.11 is A.49.10 (HWE0509), and the current patch to be applied to that is PHSS_33918.

Andrew

[Glenn Mitchell_2]

I did award points - each reply I remember putting in the points for each of the HP returns. I must not be doing something right.

The stm version I'm running is A.29.00 B4708AA.

Looking at stm I see that I still have a yellow indicator on CPU (5d3) 37 "information killed" and a grey indicator on the SCSI disk at 1/12/0/1.9.0. The ioscan of the device shows the drive is "claimed" and gpm shows all 4 CPUs as enabled with activity. Should I setup a hardware call?

[Andrew Merritt_2]

If you are running A.29.00 then you definitely need to upgrade!

There is a known problem in that release that does lead to the activity_log file filling up. The problem was fixed in the A.30.00 release for 11.11.

However, as the A.29.00 release is nearly 4 years old, you should upgrade to the A.49.10 release plus PHSS_33918. That will certainly fix the activity_log problem, as well as number of other known problems.

Andrew

[Glenn Mitchell_2]

Here is another attempt at awarding points. You all earned them because all of your advice and recommendations were good.

Thank you.

[Glenn Mitchell_2]

I will download the stm upgrades. Because this is a validated production server I will need to jump through some change control hoops before I can load it. Thank you

[Andrew Merritt_2]

Hi Glen,

From memory, you need to select the number of points for each reply, and then hit "Confirm" or something similar before the points are registered.

Andrew

[Glenn Mitchell_2]

I figured it out - there was a button to the lower left which says "Submit Points" (it took me long enough to figure out).

Do I submit a hardware call on my CPU and disk (I'm not sure there is anything wrong with them - even though stm shown a yellow and grey highlighting)?

[Andrew Merritt_2]

I'd install the upgraded OnlineDiags first, and then see if you can run the info tools on the CPU and SCSI disk. The 'Information Killed' just means that the Information tool was killed for some reason the last time it was run.

If there's actually anything wrong with the hardware there would be EMS events being generated.

Andrew