ext3 journaling

ext3 journaling

[Zachary Peterson]

I've been working on versioning extension to ext3, which requires
modifying some of the data structures, the way ext3 does writes,
truncations, etc.  But I've got a journal bug and I don't know what the
error message means, and I'm looking for some help.

Here's the output from the kernel log:

Oct 13 08:40:14 hssl1 kernel: journal_bmap: journal block not found at
offset 7948 on sd(8,1)
Oct 13 08:40:14 hssl1 kernel: Aborting journal on device sd(8,1).
Oct 13 08:40:14 hssl1 kernel: ext3_abort called.
Oct 13 08:40:14 hssl1 kernel: EXT3-fs abort (device sd(8,1)):
ext3cow_journal_start: Detected aborted journal
Oct 13 08:40:14 hssl1 kernel: Remounting filesystem read-only

Sometimes I'll get a function backtrace from the abort:

Sep 30 12:07:23 hssl1 kernel: journal_bmap: journal block not found at
offset 7948 on sd(8,1)
Sep 30 12:07:23 hssl1 kernel: Aborting journal on device sd(8,1).
Sep 30 12:07:24 hssl1 kernel: ext3_reserve_inode_write: aborting
transaction:
 Journal has aborted in __ext3_journal_get_write_access<2>EXT3COW-fs
error (device sd(8,1)) in ext3_reserve_inode_write: Journal has aborted
Sep 30 12:07:24 hssl1 kernel: EXT3-fs error (device sd(8,1)) in
ext3_new_inode: Journal has aborted
Sep 30 12:07:24 hssl1 kernel: EXT3-fs error (device sd(8,1)) in
ext3_create: Journal has aborted
Sep 30 12:07:24 hssl1 kernel: ext3_abort called.
Sep 30 12:07:24 hssl1 kernel: EXT3-fs abort (device sd(8,1)):
ext3_journal_start: Detected aborted journal
Sep 30 12:07:24 hssl1 kernel: Remounting filesystem read-only

I'm stress testing my file system by replaying 3 months of system call
traces back very rapidly.  My biggest question is what does the "journal
block not found" error mean?  Any other insights would be very helpful.

Thanks,
Zachary

=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Zachary Peterson       zachary@cse.ucsc.edu
                       http://znjp.com

856D 29FA E1F7 DB5E 9215  C68D 5F0F 3929 C929 9A72




-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
SourceForge.net hosts over 70,000 Open Source Projects.
See the people who have HELPED US provide better services:
Click here: http://sourceforge.net/supporters.php

RE: ext3 journaling

[Joseph D. Wagner]

(I'm new to fs development, so be careful with my advise!)

journal_bmap is basically a kernel-wrapper (i.e. kernel compatibility) function.

I'm willing to bet it means that you've missed changing a data structure somewhere, but since I'm unfamiliar with your project I can't really help you except to point you in the right direction.

Look in the jfs_user.h and journal.c for the journal_bmap function and all related data structures and functions.

Hope I helped.

Joseph Wagner

> -----Original Message-----
> From: Zachary Peterson [mailto:zachary@cse.ucsc.edu]
> Sent: Monday, October 13, 2003 9:49 AM
> To: linux-fsdevel@vger.kernel.org
> Cc: ext2-devel@lists.sourceforge.net
> Subject: ext3 journaling
> 
> 
> I've been working on versioning extension to ext3, which requires
> modifying some of the data structures, the way ext3 does writes,
> truncations, etc.  But I've got a journal bug and I don't know what the
> error message means, and I'm looking for some help.
> 
> Here's the output from the kernel log:
> 
> Oct 13 08:40:14 hssl1 kernel: journal_bmap: journal block not found at
> offset 7948 on sd(8,1)
> Oct 13 08:40:14 hssl1 kernel: Aborting journal on device sd(8,1).
> Oct 13 08:40:14 hssl1 kernel: ext3_abort called.
> Oct 13 08:40:14 hssl1 kernel: EXT3-fs abort (device sd(8,1)):
> ext3cow_journal_start: Detected aborted journal
> Oct 13 08:40:14 hssl1 kernel: Remounting filesystem read-only
> 
> Sometimes I'll get a function backtrace from the abort:
> 
> Sep 30 12:07:23 hssl1 kernel: journal_bmap: journal block not found at
> offset 7948 on sd(8,1)
> Sep 30 12:07:23 hssl1 kernel: Aborting journal on device sd(8,1).
> Sep 30 12:07:24 hssl1 kernel: ext3_reserve_inode_write: aborting
> transaction:
>  Journal has aborted in __ext3_journal_get_write_access<2>EXT3COW-fs
> error (device sd(8,1)) in ext3_reserve_inode_write: Journal has aborted
> Sep 30 12:07:24 hssl1 kernel: EXT3-fs error (device sd(8,1)) in
> ext3_new_inode: Journal has aborted
> Sep 30 12:07:24 hssl1 kernel: EXT3-fs error (device sd(8,1)) in
> ext3_create: Journal has aborted
> Sep 30 12:07:24 hssl1 kernel: ext3_abort called.
> Sep 30 12:07:24 hssl1 kernel: EXT3-fs abort (device sd(8,1)):
> ext3_journal_start: Detected aborted journal
> Sep 30 12:07:24 hssl1 kernel: Remounting filesystem read-only
> 
> I'm stress testing my file system by replaying 3 months of system call
> traces back very rapidly.  My biggest question is what does the "journal
> block not found" error mean?  Any other insights would be very helpful.
> 
> Thanks,
> Zachary
> 
> =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
> Zachary Peterson       zachary@cse.ucsc.edu
>                        http://znjp.com
> 
> 856D 29FA E1F7 DB5E 9215  C68D 5F0F 3929 C929 9A72
> 
> 
> -
> To unsubscribe from this list: send the line "unsubscribe linux-fsdevel"
> in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html




-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
SourceForge.net hosts over 70,000 Open Source Projects.
See the people who have HELPED US provide better services:
Click here: http://sourceforge.net/supporters.php