/var/tombstones

[Brad Marks]

I am getting approximately 3 /var/tombstones entries per month. What generates them? What should I do with them?

[Brian M. Fisher · Accepted solution]

Per document #A5738724 at:
http://us-support.external.hp.com/cki/bin/doc.pl/

"First of all, these are called tombstone files. Because you are running
diagnostics, each system boot (via 'pdcinfo') creates tombstone files.
These files contain the contents of the PIM area (non-volatile RAM). The
tombstone files provide diagnostic information if the system encounters
a hardware failure.

You can toggle 'pdcinfo' off and on in the '/etc/rc.config.d/pdcinfo'
file. This file contains one line that you can set to either 0 or 1.

Do not be concerned about the tombstone files."

Brian
<*(((>< er

[Cheryl Griffin]

The highest numbered ts# file is the latest log. Up to 100 of these files are kept. The files are created during bootup. If you are short on space, remove the oldest files first. It is best to keep a few of the most current in case there is a hardware issue that needs to be diagnosed.

[Rita C Workman]

The most important /var/tombstone file is the last one created. They are used to help you diagnose problems, especially hardware. Since you may have diagnostics turned on...your going to see these created when you boot. They become invaluable when your system panics, the last tombstone file will help you to determine if the problem was hardware or not...
Regards,

[Antoanetta Naghiu]

See what the file contains. Grep for HPMC (chassis codes). If you have them, if had a hardware error that record that tombstone. Check /var/crash for possible crash file. Call Response Center.

[Jennifer LaFene]

Here is just one more document I found which describes that all BUT ts99 can be deleted from that folder.

Per document KBRC00001750:

Current Path Home > HP-UX Software
Score
Document Type EN
Date 2000 Mar 16
Description What are the files under /var/tombstones? Can I remove them?
Document Id KBRC00001750
Search String xwindows

The processor hardware log(s) (e.g., HPMC tombstones) are copied to a file
during boot. If the system last went down with a High Priority Machine
Check (HPMC), the processor hardware failure logs will be
automatically available in the file system.

The processor hardware logs will be placed into the directory
/var/tombstones. A snapshot of the processor hardware logs of up to the
last 100 boots is kept. The files are named ts99, ts98, ..., ts0. The
highest numbered ts# file is the latest log. With that in mind, all the files
under /var/tombstones EXCEPT ts99 could be removed
to free up space in /var. These files can be removed with the rm
command.

I hope this helps.
Thank you
Jennifer-

[Brad Marks]

One of you kind folks mentioned that I should check /var/crash. I found that I don't have a /var/crash directory. Should I create one or will it be created automatically in the event of a crash?

Thanks to all!

[Brian M. Fisher]

The directory you are looking for is /var/adm/crash
This is the default directory, at least at 10.20, where system core files are placed.
The directory is set in /etc/rc.config.d/savecore with the SAVECORE_DIR variable.
To enable core file creation, in the event of a system crash, make sure SAVECORE=1 is set in the previous file.

Brian
<*(((>< er

[Cheryl Griffin]

The directory is /var/adm/crash as set as the default. You might also have this configured to be elsewhere based on SAVEDIR=
in /etc/savecore.