Re: How do you fsck a loop-aes filesystem?

[Ben Slusky <sluskyb@paranoiacs.org>]

On Wed, 03 Jul 2002 13:29:41 -0700, dave-mlist@bfnet.com wrote:
> I created a filesystem following Example 3 of the loop-aes README.  My
> /etc/fstab file has a line that looks something like this:
> 
> /dev/sda1 /mnt/crypt ext3 defaults,noauto,loop=/dev/loop0,encryption=AES128,pseed=<someseed> 0 0
> 
> So, if I wanted to run fsck on this or just bring the filesystem
> up to date with the journal, how would I do it?

I don't believe there's any easy way to do this.. here are some
(progressively more) difficult ones:

-Hack your init scripts to set up the loop device before fsck -A is run,
 then change the line in fstab to:
 /dev/loop0 /mnt/crypt ext3 defaults 1 2

-Hack your init scripts to mount /mnt/crypt read-only, fsck it, and then
 remount it read-write (all after fsck -A)

-Hack e2fsck to understand the loop= option and run losetup

> Also, is there any guarantee that the ext3 journal on this filesystem
> is being written to the physical disk?  Or is the loop device
> totally asynchronous?

I can't find the reference (one of Jari's previous posts) but ISTR that
if your loop device is block-device-backed rather than file-backed, then
journaling works as it should. I would hope so, 'cuz that's my setup as
well.

HTH,

-- 
Ben Slusky                      | "Dance like it hurts, love
sluskyb@stwing.org              |  like you need money, work
sluskyb@paranoiacs.org          |  when people are watching."
PGP keyID ADA44B3B              |               -Dogbert
-
Linux-crypto:  cryptography in and on the Linux system
Archive:       http://mail.nl.linux.org/linux-crypto/