[Patch 0/7] 2.4.20-pre4/ext3: ext3 minor improvements

[Patch 0/7] 2.4.20-pre4/ext3: ext3 minor improvements

[Stephen Tweedie]

This patch set contains a couple of minor fixups for ext3 on 2.4, plus
a couple of new checks and functionality aimed at helping InterMezzo.
It also brings the ext3 version up to the current cvs version number.

Re: [Patch 0/7] 2.4.20-pre4/ext3: ext3 minor improvements

[Marcelo Tosatti]

On Thu, 26 Sep 2002, Stephen Tweedie wrote:

> This patch set contains a couple of minor fixups for ext3 on 2.4, plus
> a couple of new checks and functionality aimed at helping InterMezzo.
> It also brings the ext3 version up to the current cvs version number.

I guess you missed the patch, Stephen ;)

Re: [Patch 0/7] 2.4.20-pre4/ext3: ext3 minor improvements

[Marcelo Tosatti]

On Fri, 27 Sep 2002, Andreas Dilger wrote:

> On Sep 27, 2002  17:44 -0300, Marcelo Tosatti wrote:
> > On Thu, 26 Sep 2002, Stephen Tweedie wrote:
> > > This patch set contains a couple of minor fixups for ext3 on 2.4, plus
> > > a couple of new checks and functionality aimed at helping InterMezzo.
> > > It also brings the ext3 version up to the current cvs version number.
> >
> > I guess you missed the patch, Stephen ;)
>
> No, they were in separate emails following this one (numbered 0/7).
> This was just the overview of all the patches.

Ok, silly: Should have read the whole message.

Re: [Patch 0/7] 2.4.20-pre4/ext3: ext3 minor improvements

[Andreas Dilger]

On Sep 27, 2002  17:44 -0300, Marcelo Tosatti wrote:
> On Thu, 26 Sep 2002, Stephen Tweedie wrote:
> > This patch set contains a couple of minor fixups for ext3 on 2.4, plus
> > a couple of new checks and functionality aimed at helping InterMezzo.
> > It also brings the ext3 version up to the current cvs version number.
> 
> I guess you missed the patch, Stephen ;)

No, they were in separate emails following this one (numbered 0/7).
This was just the overview of all the patches.

Cheers, Andreas
--
Andreas Dilger
http://www-mddsp.enel.ucalgary.ca/People/adilger/
http://sourceforge.net/projects/ext2resize/

ext3 dies without inodes

[David Fries]

[-- Attachment #1: Type: text/plain, Size: 3238 bytes --]

I'm runnig 2.4.19 and Debian (but I compile my own kernel from the
sources).  ext3 is forcing the block device to be read only when I run
out of inodes, and the only way out is reboot (that I could tell).
This is wrose than a good deal of kernel panics I've had.  Is
2.4.20prewhatever any better with reguard to this error?

My system is running low on inodes (my fault), and back when I was
using ext2 everything was fine, I would run out, the kernel would give
the message about no more space on disk for everything that needed an
inode until I freed up some.  With ext3 I get one message saying no
more space on disk and then everything else gives, readonly file
system.

After I hit the problem accidently, I verified running out of inodes
was the problem on purpose.  I would hate to think that I had to
reboot everytime I filled up my harddrive, and it is an equally bad
behavior to reboot when I run out of inodes.

Here is some dmesg output.
EXT3-fs error (device ide0(3,2)) in ext3_new_inode: error 28
Aborting journal on device ide0(3,2).
Remounting filesystem read-only
EXT3-fs error (device ide0(3,2)) in ext3_create: IO failure
EXT3-fs error (device ide0(3,2)) in start_transaction: Journal has aborted
EXT3-fs error (device ide0(3,2)) in start_transaction: Journal has aborted
ext3_abort called.
EXT3-fs abort (device ide0(3,2)): ext3_remount: Abort forced by user
ext3_abort called.
EXT3-fs abort (device ide0(3,2)): ext3_remount: Abort forced by user
ext3_abort called.
EXT3-fs abort (device ide0(3,2)): ext3_remount: Abort forced by user
ext3_abort called.
EXT3-fs abort (device ide0(3,2)): ext3_remount: Abort forced by user
EXT3-fs error (device ide0(3,2)) in start_transaction: Journal has aborted
EXT3-fs error (device ide0(3,2)) in start_transaction: Journal has aborted
EXT3-fs error (device ide0(3,2)) in start_transaction: Journal has aborted
EXT3-fs error (device ide0(3,2)) in start_transaction: Journal has aborted
EXT3-fs error (device ide0(3,2)) in start_transaction: Journal has aborted
EXT3-fs error (device ide0(3,2)) in start_transaction: Journal has aborted
EXT3-fs error (device ide0(3,2)) in start_transaction: Journal has aborted
EXT3-fs error (device ide0(3,2)) in start_transaction: Journal has aborted
EXT3-fs error (device ide0(3,2)) in start_transaction: Journal has aborted
EXT3-fs error (device ide0(3,2)) in start_transaction: Journal has aborted
EXT3-fs error (device ide0(3,2)) in start_transaction: Journal has aborted
EXT3-fs error (device ide0(3,2)) in start_transaction: Journal has aborted
EXT3-fs error (device ide0(3,2)) in start_transaction: Journal has aborted
EXT3-fs error (device ide0(3,2)) in start_transaction: Journal has aborted
ext3_abort called.
EXT3-fs abort (device ide0(3,2)): ext3_remount: Abort forced by user
ext3_abort called.
EXT3-fs abort (device ide0(3,2)): ext3_remount: Abort forced by user
SpacedOut:/mnt/david/tuxscreen/buildroot-tux/build$ df -i
Filesystem            Inodes   IUsed   IFree IUse% Mounted on
/dev/hda2             362304  361710     594  100% /
/dev/hdb1            1664640  210988 1453652   13% /mnt/hdb1


-- 
David Fries <dfries@mail.win.org>
http://fries.net/~david/pgpkey.txt

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: ext3 dies without inodes

[Andreas Dilger]

On Oct 29, 2002  13:00 -0600, David Fries wrote:
> I'm runnig 2.4.19 and Debian (but I compile my own kernel from the
> sources).  ext3 is forcing the block device to be read only when I run
> out of inodes, and the only way out is reboot (that I could tell).
> This is wrose than a good deal of kernel panics I've had.  Is
> 2.4.20prewhatever any better with reguard to this error?

Yes, this is fixed in more recent kernels.  Separate patches are also
available if you want to stick with 2.4.19 for whatever reason.
Stephen posted a URL for them a couple of times.

Cheers, Andreas
--
Andreas Dilger
http://www-mddsp.enel.ucalgary.ca/People/adilger/
http://sourceforge.net/projects/ext2resize/

Re: ext3 dies without inodes

[Stephen Tweedie]

Hi,

On Tue, Oct 29, 2002 at 12:14:53PM -0700, Andreas Dilger <adilger@clusterfs.com> wrote:
> On Oct 29, 2002  13:00 -0600, David Fries wrote:
> > I'm runnig 2.4.19 and Debian (but I compile my own kernel from the
> > sources).  ext3 is forcing the block device to be read only when I run
> > out of inodes, and the only way out is reboot (that I could tell).
> > This is wrose than a good deal of kernel panics I've had.  Is
> > 2.4.20prewhatever any better with reguard to this error?
 
> Yes, this is fixed in more recent kernels.  Separate patches are also
> available if you want to stick with 2.4.19 for whatever reason.
> Stephen posted a URL for them a couple of times.

   http://people.redhat.com/sct/patches/ext3-2.4/for-2.4.19/all-in-one.patch

Cheers,
 Stephen