Re: Ext3 Errors...

[Jan Kara]

  Hi,

> I have something really strange while running kernel 2.6.19.
> 
> See the following lines:
> 
> Dec  5 23:50:49 kernel: do_get_write_access: OOM for frozen_buffer
> Dec  5 23:50:49 kernel: ext3_free_blocks_sb: aborting transaction: Out of
> memory in __ext3_journal_get_undo_access
> Dec  5 23:50:49 kernel: EXT3-fs error (device md1) in ext3_free_blocks_sb:
> Out of memory
> Dec  5 23:50:49 kernel: EXT3-fs error (device md1) in
> ext3_reserve_inode_write: Readonly filesystem
> Dec  5 23:50:50 kernel: EXT3-fs error (device md1) in ext3_truncate: Out
> of memory
> Dec  5 23:50:51 kernel: EXT3-fs error (device md1) in
> ext3_reserve_inode_write: Readonly filesystem
> Dec  5 23:50:51 kernel: EXT3-fs error (device md1) in ext3_orphan_del:
> Readonly filesystem
> Dec  5 23:50:51 kernel: EXT3-fs error (device md1) in
> ext3_reserve_inode_write: Readonly filesystem
> Dec  5 23:50:51 kernel: EXT3-fs error (device md1) in ext3_delete_inode:
> Out of memory
> 
> And three days later the same:
> 
> Dec  8 08:24:29 kernel: do_get_write_access: OOM for frozen_buffer
> Dec  8 08:24:29 kernel: ext3_reserve_inode_write: aborting transaction:
> Out of memory in __ext3_journal_get_write_access
> Dec  8 08:24:29 kernel: EXT3-fs error (device md1) in
> ext3_reserve_inode_write: Out of memory
> Dec  8 08:24:32 kernel: EXT3-fs error (device md1) in ext3_dirty_inode:
> Out of memory
> Dec  8 08:24:32 kernel: EXT3-fs error (device md1) in ext3_new_blocks:
> Readonly filesystem
> Dec  8 08:24:32 kernel: EXT3-fs error (device md1) in
> ext3_reserve_inode_write: Readonly filesystem
> Dec  8 08:24:32 kernel: EXT3-fs error (device md1) in ext3_dirty_inode:
> Out of memory
> Dec  8 08:24:32 kernel: EXT3-fs error (device md1) in ext3_prepare_write:
> Out of memory
> 
> Now the funny thing is, with kernel 2.6.18.3 I did not had these errors.
> Could it be my memory that is just going nuts, or something else? I have
> seen some other topics about the EXT3 corruption problems. Maybe this is
> also the same thing?
  Probably some error on the kernel's side. Maybe somebody (e.g. my new
code in jbd commit) forgets to release some buffers? What does your
/proc/slabinfo look like? Thanks for report.

								Honza
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/