Re: I think I just got a panic

[Matthew Dillon <dillon@xxxxxxxxxxxxxxxxxxxx>]

:But I was running X at the time :( X froze and stayed there for about 10 
:seconds or something, then the computer rebooted. Probably a panic?
:
:For whatever it's worth I was just doing a buildworld, X with a couple 
:of xterms open (icewm) and I had a samba share mounted. I was ssh'd into 
:another machine in one of the xterms and was about to vi /etc/gettytab 
:locally when it locked up and rebooted.
:
:I've now set dumpdev in rc.conf. Is that enough to atleast save a dump, 
:so I can tell for sure if it was a panic or not, incase it happens again?
:-- 

    It sounds like it auto-rebooted after a crash.  The 10 seconds was
    probably the 'will reboot after 15 seconds' message on the console
    which you didn't see because you were in X.  With your dumpdev set if
    you have another crash you should see the disk drive LED go solid for
    a while, indicating that it is writing out the crash dump to disk. 
    It may take a little longer before it reboots, depending on how much
    memory you have (that it has to copy to the disk), so if the disk drive
    LED is on be a patient and you should have a crash dump when you come
    up again.

    Note that the system will attempt to store the crash dump in /var/crash.
    If /var/crash does not have sufficient space, the system will not be 
    able to save a crash dump.  What most people do is:

	cpdup /var/crash /usr/var.crash
	rm -rf /var/crash
	ln -s /usr/var.crash /var/crash

    or

	cpdup /var/crash /home/var.crash
	rm -rf /var/crash
	ln -s /home/var.crash /var/crash

    So /var/crash has enough space.

    That's actually something we should fix in the installer... to make sure
    that /var has enough space to hold crash dumps (a nominal ~2-3x the space
    would be good if it doesn't eat too much from other partitions.  i.e. if
    the disk is big enough).

					-Matt
					Matthew Dillon 
					<dillon@xxxxxxxxxxxxx>