I have a system which had bad memory.
Bad memory was replaced, but filesystem corruption caused by it (and related crashes) is still there:
[104472.909769] EXT4-fs (vdb1): mounted filesystem with ordered data mode. Opts: errors=remount-ro
[104773.600088] EXT4-fs (vdb1): error count: 2
[104773.600093] EXT4-fs (vdb1): initial error at 1290468286: htree_dirblock_to_tree:586: inode 30204507: block 120625971
[104773.600097] EXT4-fs (vdb1): last error at 1290468334: ext4_put_super:726: inode 30204507: block 120625971
Unfortunately, repeated fsck checks do not make the error go away (using 2.6.37-rc3 and e2fsprogs 1.41.12):
# e2fsck -f -v -y /dev/vdb1
e2fsck 1.41.12 (17-May-2010)
Pass 1: Checking inodes, blocks, and sizes
Pass 2: Checking directory structure
Pass 3: Checking directory connectivity
Pass 4: Checking reference counts
Pass 5: Checking group summary information
16281559 inodes used (35.49%)
4484 non-contiguous files (0.0%)
29082 non-contiguous directories (0.2%)
# of inodes with ind/dind/tind blocks: 0/0/0
Extent depth histogram: 16273127/8421
152806601 blocks used (83.27%)
0 bad blocks
27 large files
7931312 regular files
8350237 directories
0 character device files
0 block device files
0 fifos
95162835 links
0 symbolic links (0 fast symbolic links)
1 socket
--------
111444385 files
# mount -a
# dmesg -c
[108692.626550] EXT4-fs (vdb1): mounted filesystem with ordered data mode. Opts: errors=remount-ro
# ls -lR /mount/point
(...)
# dmesg -c
[108993.760041] EXT4-fs (vdb1): error count: 2
[108993.760051] EXT4-fs (vdb1): initial error at 1290468286: htree_dirblock_to_tree:586: inode 30204507: block 120625971
[108993.760058] EXT4-fs (vdb1): last error at 1290468334: ext4_put_super:726: inode 30204507: block 120625971
The error is exactly the same as the first one I pasted (multiple umount, fsck, mount runs without reboot).
Is this all something to be worry about?
--
Tomasz Chmielewski
http://wpkg.org
On 2010-11-25, at 07:57, Tomasz Chmielewski wrote:
> I have a system which had bad memory.
>
> Bad memory was replaced, but filesystem corruption caused by it (and related crashes) is still there:
>
> [104472.909769] EXT4-fs (vdb1): mounted filesystem with ordered data mode. Opts: errors=remount-ro
> [104773.600088] EXT4-fs (vdb1): error count: 2
> [104773.600093] EXT4-fs (vdb1): initial error at 1290468286: htree_dirblock_to_tree:586: inode 30204507: block 120625971
> [104773.600097] EXT4-fs (vdb1): last error at 1290468334: ext4_put_super:726: inode 30204507: block 120625971
The "initial error" handling has been fixed in the e2fsprogs git repo. Search the archives for the patch.
> The error is exactly the same as the first one I pasted (multiple umount, fsck, mount runs without reboot).
>
> Is this all something to be worry about?
No, this is just a historical log of what caused the filesystem problems. It doesn't mean that there is anything wrong with the filesystem now.
Cheers, Andreas