Re: fsck on boot...revisited
----- Original Message -----
> On Thu, Jul 25, 2013 at 10:09:00AM -0500, Tim Nelson wrote:
> > ----- Original Message -----
> > > On Thu, Jul 25, 2013 at 09:28:56AM -0500, Tim Nelson wrote:
> > > > I have an interesting use case where a Debian Lenny server runs
> > > > headless, and is at the mercy of poor power conditions
> > > > (environmental monitoring at a remote storage building). We
> > > > used
> > > > to have issues with the server not coming up after several
> > > > reboots, but we gave it a bandaid by forcing an fsck on every
> > > > boot
> > > > (tune2fs...) to correct any issues. This is fine, and has done
> > > > wonders for disk errors.
> > > >
> > > > However...
> > > >
> > > > On occasion, we find that a filesystem error is bad enough that
> > > > instead of auto{matically|magically} fixing the issue and
> > > > continuing to boot, the system hangs, needing a root password
> > > > entered for a manual fsck to be run.
> > > >
> > > > My question is thus: How do I prevent that requirement to login
> > > > and
> > > > run fsck manually? Is there some parameter that can be set? Or,
> > > > am
> > > > I going about this the completely wrong way?
> > >
> > > Solve the underlying problem as best you can.
> > >
> > > Buy a cheap UPS with a serial or USB interface; run the
> > > appropriate daemon on your server to shut it down automatically
> > > when the power drops. Replace the UPS every year or two.
> > >
> > > Now your disks will be much happier.
> > >
> >
> > Yes, that is the *proper* solution. But, how about a situation
> > where I want to solve this even if battery power goes out?
> >
> > Does the 'FSCKFIX' option within /etc/default/rcS do what I need?
>
> Yes, but your disks will continue to degrade. One morning you
> will wake up to find a dead server, much earlier than you would
> otherwise.
>
How would they continue to degrade? Meaning, what is the difference between the system answering 'YES' to the manual prompts versus someone driving out and doing the same thing? Of course, we're going on the premise that the power situation cannot be fixed, and the solution is to get the best reliability out of the system when it experiences somewhat random/often reboots. I know it isn't ideal...
--Tim
Reply to: