On Tue, Jun 29, 2010 at 04:30:42PM +0200, John Hughes wrote:
> Package: e2fsprogs
> Version: 1.41.12-2
> Severity: normal
>
> On a lvm2 based system I have an ext3 fileystem with an external journal.
>
> I want to use lvm2s snapshot facilities to do online backups and fscks.
>
> So I create a snapshot of my filesystem.
>
> Problem #1 - if I try to fsck the snapshot it proposes that I rerun the
> journal, even suggesting that the journal be cleared.
>
> # fsck -v -f /dev/group-a/ums-snap
> fsck from util-linux-ng 2.16.2
> e2fsck 1.41.12 (17-May-2010)
> Superblock needs_recovery flag is clear, but journal has data.
> Run journal anyway<y>? no
>
> Clear journal<y>? no
>
> If I do rerun the journal the snapshot is, of course, corrupted. I
> imagine that if I clear the journal the filesystem will be damaged.
Yeah, that's a real problem. Right now e2sprogs doesn't really know
about LVM snapshots in any kind of fundamental way. So it doesn't
know that it's checking a snapshot, and so ignoring the journal is the
right thing to do.
I need a way to determine reliably and portably, whether or not a
volume is a snapshot or not. I'm not sure what's the best way to do
it right now, but I wanted to ack this as an accepted bug.
The short-term fix which I could offer is a command-line option to
force e2fsck to ignore the journal, but that really isn't the best way
to go, since users will often forget to give this option, and the
results would then be unfortunate. I suppose it might be a sufficient
if the e2fsck is being invoked from a shell script, but I would like
to see if I can do better...
Regards,
- Ted
--
To UNSUBSCRIBE, email to [email protected]
with a subject of "unsubscribe". Trouble? Contact [email protected]
