On Wed, 23 Jun 2004 11:44:47 -0700, Steve Simitzis wrote:

> heh. :)
> 
> sounds like the same dance i had to do a couple of months ago.
> 
> coda spits out information into its logs before it crashes. look for
> errors in the last couple of lines of /usr/coda/etc/console and possibly
> /usr/coda/etc/venus.log

console (it's called venus.err on debian) told me the usual
crashed-getting zombie- use gdb thing. I didn't check venus.log at that
time, will do the next time it happens. I've since lowered the venus cache
size from 8Gb to 6Gb, the last two reboots with full cache went smooth.

> you can also run gdb -p [pid] on the zombie venus process. in my venus
> installation, it doesn't actually leave a core file hanging around, but
> you can run gdb on the same lingering venus process.

I don't have any gdb experience so far, what things should i look out for?

> please post anything you're able to dig out to the list. i would be
> curious to know if you're getting the same type of crashing that i've been
> getting.

Will do.

-- 
Gabriel Wicke