On 2013/01/01 11:59 AM, "Brock Palen" <brockp(a)umich.edu> wrote:
I am fscking a lustre OST lun, and it shows the following:
[root@oss4 ~]# e2fsck -f -C 0 /dev/mapper/mpath7
e2fsck 1.42.3.wc1 (28-May-2012)
Pass 1: Checking inodes, blocks, and sizes
Inode 192820368, i_blocks is 15512, should be 4294982808. Fix? yes
Pass 2: Checking directory structure
i_blocks_hi for inode 192820368 (/O/0/d0/42065024) is 1, should be zero.
Clear? yes
Pass 3: Checking directory connectivity
Pass 4: Checking reference counts
Pass 5: Checking group summary information
nobackup-OST001b: ***** FILE SYSTEM WAS MODIFIED *****
nobackup-OST001b: 1756205/477626368 files (5.6% non-contiguous),
1529484010/1910505472 blocks
I have run fsck twice and the same error with that same inode appeared
both times and worries me, why when I say to clear it it doesn't really
clear. Any thought on this?
This looks like a generic e2fsck bug. Is the inode actually larger than
16TB, or does it just look that way?
Cheers, Andreas
--
Andreas Dilger
Lustre Software Architect
Intel High Performance Data Division