2008-04-22 00:18:59

by Badari Pulavarty

[permalink] [raw]
Subject: ext4 errors while runing fsx (2.6.25)

Got these messages while running fsx on ext4.
Known issue ?

Thanks,
Badari

EXT4 FS on hda1, internal journal
EXT4-fs: mounted filesystem with ordered data mode.
EXT4-fs: file extents enabled
EXT4-fs: mballoc enabled
pa ffff8101cc8400b0: logic 8, phys. 9065, len 56
EXT4-fs error (device hda1): ext4_mb_release_inode_pa: free 37, pa_free 38

pa ffff8101cc8556b0: logic 32, phys. 8935, len 32
EXT4-fs error (device hda1): ext4_mb_release_inode_pa: free 21, pa_free 22

pa ffff8101cc85e8b0: logic 9, phys. 8971, len 55
EXT4-fs error (device hda1): ext4_mb_release_inode_pa: free 43, pa_free 44

pa ffff81017d0f7330: logic 39, phys. 9131, len 25
EXT4-fs error (device hda1): ext4_mb_release_inode_pa: free 20, pa_free 21

pa ffff8101bb1431b0: logic 18, phys. 33438, len 46
EXT4-fs error (device hda1): ext4_mb_release_inode_pa: free 26, pa_free 27




2008-04-23 08:45:44

by Aneesh Kumar K.V

[permalink] [raw]
Subject: Re: ext4 errors while runing fsx (2.6.25)

On Mon, Apr 21, 2008 at 05:18:53PM -0700, Badari Pulavarty wrote:
> Got these messages while running fsx on ext4.
> Known issue ?
>
> Thanks,
> Badari
>
> EXT4 FS on hda1, internal journal
> EXT4-fs: mounted filesystem with ordered data mode.
> EXT4-fs: file extents enabled
> EXT4-fs: mballoc enabled
> pa ffff8101cc8400b0: logic 8, phys. 9065, len 56
> EXT4-fs error (device hda1): ext4_mb_release_inode_pa: free 37, pa_free 38
>
> pa ffff8101cc8556b0: logic 32, phys. 8935, len 32
> EXT4-fs error (device hda1): ext4_mb_release_inode_pa: free 21, pa_free 22
>
> pa ffff8101cc85e8b0: logic 9, phys. 8971, len 55
> EXT4-fs error (device hda1): ext4_mb_release_inode_pa: free 43, pa_free 44
>
> pa ffff81017d0f7330: logic 39, phys. 9131, len 25
> EXT4-fs error (device hda1): ext4_mb_release_inode_pa: free 20, pa_free 21
>
> pa ffff8101bb1431b0: logic 18, phys. 33438, len 46
> EXT4-fs error (device hda1): ext4_mb_release_inode_pa: free 26, pa_free 27

One thing I observed is free and pa_free are all differing by 1. I
looked at the code and we seems to be counting right. Can I get access
to the machine where it happened ?. I would like to look at the bitmap.

-aneesh