crash

[Sajjad Sahir]

how to check the system crash information?

[Matt Hearn · Accepted solution]

There are couple places to check. If you have enough space in /var or /var/adm/crash (if it's a separate filesystem) to save the entire memory contents, you can look in /var/adm/crash for any recent entries. There's also /var/tombstones. However a lot of this stuff is really only useful if you have the tools to decode it (HP does, so your best bet is to open a ticket and send them the files).

If you're lucky, /var/adm/syslog/OLDsyslog.log will have some useful info; look at the end of it for obvious hardware problems or OS panics.

Depending on your hardware model, you might be able to check the MP/console logs for hardware alerts (High Priority Machine Checks, or HPMCs, and the like).

[Michael Steele_2]

The command 'crashconf' will list your current dump configuration.

The command 'savecrash -rf /dir/file' file try to save any missed crash dump data from the command line. But chances are this file will be incomplete as the crash data will be partially or totally overwritten during boot up. This is because swap and dump share the same logical volume, lvol2 in vg00.

The command 'lvlnboot' -d or 'lvlnboot -v' will verify that dump has been properly configured in lvol2.

There's a lot more to configuring dump including the /etc/rc.config.d/crashconf file to point to /var/adm/crash. This is the file read by the 'crashconf' command.

Refer to this link for more information or search the ITRC using 'crashconf':

http://forums12.itrc.hp.com/service/forums/questionanswer.do?threadId=112398&admit=109447627+1202577251067+28353475

[Sajjad Sahir]

closed