Filesystem check

[Lamont R. Peterson]

On Thursday 04 May 2006 02:49am, Emeric Maschino wrote:
> > Has something changed in the manner the filesystems are checked at boot
> > time?
> >
> > Indeed, at startup, once the logical volumes are configured and
> > activated, my filesystem is checked but the check fails with the
> > following message:
> >
> > e2fsck: Cannot continue, aborting.
> >
> > I'm then asked to enter the root password in order to manually check the
> > filesystem. At this stage, the filesystem is already mounted, so I
> > pressed Ctrl-D to continue and my system was rebooted.
> >
> > The strange thing is that, if I boot using a rescue CD in order to run
> > e2fsck on unmounted filesystems, everything is clean.
> >
> > Yesterday, I had the bad idea to finally enter the root password and
> > manually run fsck as requested. Of course, this ended up with a totally
> > corrupted filesystem. Fortunately, my data are on a separate disk. I've
> > then performed a network install of Fedora Core Rawhide during this
> > night (most recent packages dated May 2nd). Everything ran fine, but at
> > boot time, e2fsck is still failing and ask for a manual intervention!
> > Rebooting the system (Windows inside (R)) has no effect.
> >
> > Any hint ? This is on an Itanium workstation (ia64) and I don't know if
> > this problem also appears on other architectures.
>
> Removing the rhgb quiet boot options as suggested by Paul allowed me to
> log in (but you need to cancel the filesystem check when asked for the
> root password). As stated by Bill Nottingham, updating to the latest
> mkinitrd and rebuilding the initrd image definitely solved the problem
> (the rhgb quiet boot options can be passed again to the kernel).

rhgb and quiet are two separate options.  They do not have anything to do with 
each other.

IIRC, quite is understood by the kernel and causes it to not spew lots-o-text 
about it's initialization.  You see the audit message and then the next text 
("Welcome to Fedora Core" or however it's phrased) comes 
from /etc/rc.sysinit .

The rhgb option is not understood by the kernel and is, therefore, passwd on 
to init, which is patched to understand that the rhgb option should be passed 
on (in some way) to rc.sysinit which then launches the X server and the rhgb 
display program.

Are those details all correct?

Anyway, since quiet and rhgb are separate items, I wonder which one actually 
had the effect on your booting.  Or was it from interactions coming from 
both?  Not that it's a big deal, since mkinitrd is now fixed.
-- 
Lamont R. Peterson <lamont at gurulabs.com>
Senior Instructor
Guru Labs, L.C. [ http://www.GuruLabs.com/ ]
GPG Key fingerprint: F98C E31A 5C4C 834A BCAB  8CB3 F980 6C97 DC0D D409
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
URL: <http://listman.redhat.com/archives/fedora-devel-list/attachments/20060504/1bef17a2/attachment.sig>