Eu tenho um NAS qnap que está executando uma versão do Ubuntu e não estou recebendo nenhuma ajuda do fórum deles, então eu pensei em fazer minha pergunta aqui.
Eu "tinha" 3 HDs de 2 TB em uma configuração de RAID 5. Quando voltei das minhas férias de férias, descobri que o volume do ataque não estava montado. Eu corri o utilitário SMART da página de administração da QNAP e detectei erros de leitura no Disk2, então comprei outro disco de 2 TB, fiz uma troca e parecia que ia consertar o meu problema:
[~] # cat /proc/mdstat
Personalities : [linear] [raid0] [raid1] [raid10] [raid6] [raid5] [raid4] [multipath]
md0 : active raid5 sdd3[3] sdc3[2] sda3[0]
3903891200 blocks level 5, 64k chunk, algorithm 2 [3/2] [U_U]
[========>............] recovery = 41.1% (802441984/1951945600) finish=285.7min speed=67047K/sec
Eu fui para a cama e deixei terminar a reconstrução. Na manhã seguinte, tentei criar o array e ele falhou:
[~] # mdadm -CfR --assume-clean /dev/md0 -l 5 -n 3 /dev/sdb3 /dev/sda3 /dev/sdc3
mdadm: /dev/sdb3 appears to contain an ext2fs file system
size=-391076224K mtime=Thu Jan 2 03:30:10 2014
mdadm: /dev/sdb3 appears to be part of a raid array:
level=raid5 devices=3 ctime=Tue Jan 7 10:01:03 2014
mdadm: /dev/sda3 appears to contain an ext2fs file system
size=-403668988K mtime=Sat Jan 15 17:45:38 2011
mdadm: /dev/sda3 appears to be part of a raid array:
level=raid5 devices=3 ctime=Tue Jan 7 10:01:03 2014
mdadm: /dev/sdc3 appears to be part of a raid array:
level=raid5 devices=3 ctime=Tue Jan 7 10:01:03 2014
mdadm: array /dev/md0 started.
[~] # mount /dev/md0 /share/MD0_DATA -t ext4
mount: wrong fs type, bad option, bad superblock on /dev/md0,
missing codepage or other error
In some cases useful info is found in syslog - try
dmesg | tail or so
[~] # dmesg | tail
[ 5474.498500] EXT4-fs (md0): ext4_check_descriptors: Block bitmap for group 1952 not in group (block 224078240)!
[ 5474.504503] EXT4-fs (md0): group descriptors corrupted!
Corri o e2fsck e ele reclamou do superbloco:
[~] # e2fsck -f /dev/md0
e2fsck 1.41.4 (27-Jan-2009)
e2fsck: Group descriptors look bad... trying backup blocks...
e2fsck: Filesystem revision too high while trying to open /dev/md0
The filesystem revision is apparently too high for this version of e2fsck.
(Or the filesystem superblock is corrupt)
The superblock could not be read or does not describe a correct ext2
filesystem. If the device is valid and it really contains an ext2
filesystem (and not swap or ufs or something else), then the superblock
is corrupt, and you might try running e2fsck with an alternate superblock:
e2fsck -b 8193 <device>
Eu tentei superblocos alternativos:
[~] # dumpe2fs /dev/md0 | grep -i superblock
dumpe2fs 1.41.4 (27-Jan-2009)
Primary superblock at 0, Group descriptors at 1-233
Backup superblock at 32768, Group descriptors at 32769-33001
Backup superblock at 98304, Group descriptors at 98305-98537
Backup superblock at 163840, Group descriptors at 163841-164073
...
Quando eu experimentei o segundo valor, ele cuspiu todos os tipos de erros, aqui está um trecho deles:
[~] # e2fsck -b 98304 /dev/md0
e2fsck 1.41.4 (27-Jan-2009)
Superblock has an invalid journal (inode 8).
answer=1
*** ext3 journal has been deleted - filesystem is now ext2 only ***
Block bitmap for group 1920 is not in group. (block 223029632)
answer=1
Inode bitmap for group 1920 is not in group. (block 971776033)
answer=1
Inode table for group 1920 is not in group. (block 63045632)
WARNING: SEVERE DATA LOSS POSSIBLE.
answer=1
Group descriptor 1920 marked uninitialized without feature set.
answer=1
Block bitmap for group 1921 is not in group. (block 62916993)
answer=1
Inode bitmap for group 1921 is not in group. (block 62916994)
answer=1
Group descriptor 1921 marked uninitialized without feature set.
answer=1
Block bitmap for group 1922 is not in group. (block 1236358788)
answer=1
Group descriptor 1922 marked uninitialized without feature set.
answer=1
Inode table for group 1923 is not in group. (block 2996300427)
WARNING: SEVERE DATA LOSS POSSIBLE.
...
Resize inode not valid. answer=1
/dev/md0 contains a file system with errors, check forced.
Pass 1: Checking inodes, blocks, and sizes
Root inode is not a directory. answer=1
Inode 241 has EXTENTS_FL flag set on filesystem without extents support.
answer=1
Inode 263, i_blocks is 139224, should be 139056. answer=1
Inode 264 has illegal block(s). answer=1
Illegal block #2060 (2770694604) in inode 264. answer=1
CLEARED.
Illegal block #2061 (2431456604) in inode 264. answer=1
CLEARED.
...
Too many illegal blocks in inode 264.
answer=1
Inodes that were part of a corrupted orphan linked list found. answer=1
Inode 8433 was part of the orphaned inode list. answer=1
FIXED.
Inode 8433 has imagic flag set. answer=1
Inode 8433 has a extra size (48936) which is invalid
answer=1
Inode 8434 was part of the orphaned inode list. answer=1
FIXED.
Inode 8434 has imagic flag set. answer=1
Inode 8434 has a extra size (49000) which is invalid
...
Inode 16369 is in use, but has dtime set. answer=1
Inode 16369 has a extra size (35995) which is invalid
answer=1
Inode 16370 is in use, but has dtime set. answer=1
Inode 16370 has imagic flag set. answer=1
Inode 16370 has a extra size (26087) which is invalid
answer=1
Inode 16371 is in use, but has dtime set. answer=1
Inode 16371 has imagic flag set. answer=1
Inode 16371 has a extra size (1476) which is invalid
answer=1
Inode 16372 has EXTENTS_FL flag set on filesystem without extents support.
...
Inode 16380 has EXTENTS_FL flag set on filesystem without extents support.
answer=1
Inode 16381 has EXTENTS_FL flag set on filesystem without extents support.
answer=1
Inode 16382 has EXTENTS_FL flag set on filesystem without extents support.
answer=1
Inode 16383 has EXTENTS_FL flag set on filesystem without extents support.
answer=1
Inode 16384 has EXTENTS_FL flag set on filesystem without extents support.
answer=1
Inode 16375 has compression flag set on filesystem without compression support. answer=1
Inode 16375 has a bad extended attribute block 335544554. answer=1
Inode 16375 has INDEX_FL flag set but is not a directory.
answer=1
Inode 16375, i_size is 4795354609508463505, should be 0. answer=1
Inode 16375, i_blocks is 119976829646520, should be 0. answer=1
Inode 16373 has compression flag set on filesystem without compression support. answer=1
...
Inode 49141 is in use, but has dtime set. answer=1
Inode 49141 has a extra size (53130) which is invalid
answer=1
Inode 49142 is in use, but has dtime set. answer=1
Inode 49142 has a extra size (15011) which is invalid
answer=1
Inode 49143 has EXTENTS_FL flag set on filesystem without extents support.
answer=1
Inode 49144 has EXTENTS_FL flag set on filesystem without extents support.
...
Eu finalmente fiz um ctrl-C para parar, com medo de causar mais problemas. Em seguida, ele cuspiu isso:
^CRecreate journalanswer=1
Creating journal (32768 blocks):
Done.
*** journal has been re-created - filesystem is now ext3 again ***
/dev/md0: e2fsck canceled.
/dev/md0: ***** FILE SYSTEM WAS MODIFIED *****
[~] #
Qualquer ajuda sobre o que tentar em seguida será apreciada. Eu tive algumas perguntas sobre o raid e mdadm desde que eu sou um novato quando se trata deles.
o que acontece se você tentar montá-los fora de ordem, ou seja,
mdadm -CfR --assume-clean / dev / md0 -l 5 -n 3 / dev / sda3 / dev / sdb3 / dev / sdc3
em vez de
mdadm -CfR --assume-clean / dev / md0 -l 5 -n 3 / dev / sdb3 / dev / sda3 / dev / sdc3
e por que, quando executo o comando mdadm, minha nova unidade "sdb3" é a única que mostra a mensagem: parece conter um sistema de arquivos ext2fs ? Isso poderia ser parte do meu problema?
Obrigado Robert
Criar uma matriz, como formatar uma partição, é destrutivo; você não faz isso onde você já tem, a menos que queira apagá-lo e começar de novo.
Desde que você fez --assume-clean, esperamos que não tenha havido muito dano. Você provavelmente entendeu errado a ordem dos discos quando recriou a matriz, portanto, será necessário tentar novamente com a ordem correta.