Hello all!
I have a "big" filesystem (8TB), and a BIG problem!
I can't fix it!
[root@dy-base xfsprogs-2.5.6]# xfs_check /dev/md31
xfs_check: out of memory
[root@dy-base xfsprogs-2.5.6]# free
total used free shared buffers cached
Mem: 4149468 4015820 133648 0 16 3718360
-/+ buffers/cache: 297444 3852024
Swap: 0 0 0
The xfs_check always do this, and I don't use it....
But, in this time, the xfs_repair is fails ALWAYS too!
I have tryed it 23 times!
The different results:
1. Simple segfault.
2. xfs_repair: dir2.c:1471: process_dir2_data: Assertion
`(((irec_p)->ino_confirmed & ((__uint64_t)1 << ((ino_off)))) == 0LL ? 0: 1)'
failed.
3.
rebuilding directory inode 4211081386
xfs_repair: phase6.c:2604: longform_dir2_rebuild_finish: Assertion `error == 0'
failed.
4.
xfs_repair: dino_chunks.c:594: process_inode_chunk: Assertion
`((first_irec->ino_startnum) & ((__uint32_t)((1ULL <<
(((mp)->m_sb.sb_inopblog))) - 1))) == 0' failed.
5.
- traversing filesystem starting at / ...
xfs_repair: incore_ino.c:609: add_inode_reached: Assertion
`ino_rec->ino_un.backptrs != ((void *)0)' failed.
6.
- agno = 254
xfs_repair: dir2.c:1471: process_dir2_data: Assertion
`(((irec_p)->ino_confirmed & ((__uint64_t)1 << ((ino_off)))) == 0LL ? 0: 1)'
failed.
7.
- agno = 254
xfs_repair: dir2.c:939: process_sf_dir2: Assertion `(((irec_p)->ino_confirmed &
((__uint64_t)1 << ((ino_off)))) == 0LL ? 0 : 1)' failed.
The data on fs and the service is VERY important for me.
The all logs is there: http://download.netcenter.hu/xfs/xfs.log.tar.gz (215kb)
But unfortunately the first log is missing.
Can you help me?
The system:
Dual xeon, 4G ram.
The xfs is on RAID 0, 32k chunks
OS: RH9.0, kernel 2.6.13-rc3
xfsprogs: xfsprogs-2.5.6-1.i386.rpm
(sorry for my english)
Thanks a lot!
Janos
[[HTML alternate version deleted]]
|