Subject: Re: filesystem code should not panic
To: Chuck Silvers <chuq@chuq.com>
From: Chris G. Demetriou <cgd@netbsd.org>
List: tech-kern
Date: 11/23/1999 15:11:38
Chuck Silvers <chuq@chuq.com> writes:
> On Tue, Nov 23, 1999 at 12:23:30PM -0800, Chris G. Demetriou wrote:
> > Manuel Bouyer <bouyer@antioche.lip6.fr> writes:
> > > Aggred for user mounts but I want to keep the possibility for root mounts
> > > to stop the machine when a FS goes bad. Keeping the machine running without an
> > > important FS can have bad consequences (think mail servers ...).
> > 
> > There are many situations in which continuing to run is a bad thing.
> 
> and I do agree with that, I just don't think this is one of those situations.
> disk corruption on disk isn't something that a reboot persay will fix.

Right.  I agree.  A user, however, might want to reboot.  They might
want to page somebody.  They might want ... lots of things.  8-)

In general, the right way to handle this is _not_ to panic, but to
give something else (e.g. a watchdog or other system monitoring
daemon) the chance to do whatever the local administrator(s) think is
appropriate.


cgd
-- 
Chris Demetriou - cgd@netbsd.org - http://www.netbsd.org/People/Pages/cgd.html
Disclaimer: Not speaking for NetBSD, just expressing my own opinion.