怎样恢复 fsck 后损坏的 ext4 文件系统?

怎样恢复 fsck 后损坏的 ext4 文件系统?

我在 luks 上有一个 ext4 文件系统,基于软件 raid5。几年来,当我开始用完空间时,文件系统一直运行良好。我在 6x2T 驱动器上有一个 9T 卷。我开始升级到 3T 驱动器,方法是执行 mdadm 失败、删除、添加、重建、重复过程,直到我拥有更大的阵列。然后我扩大了 luks 容器,然后当我卸载并尝试 resize2fs 时,我收到消息称文件系统很脏,需要 e2fsck。

我没有多想,就直接执行了 e2fsck -y /dev/mapper/candybox,然后它开始发出各种 inode 被删除类型的消息(记不清具体是什么了),我关闭了 e2fsck,并尝试重新挂载文件系统以备份我所关心的数据。此时尝试挂载时,我得到:

# mount /dev/mapper/candybox /candybox
mount: wrong fs type, bad option, bad superblock on /dev/mapper/candybox,
       missing codepage or helper program, or other error
       In some cases useful info is found in syslog - try
       dmesg | tail  or so

回顾我的旧日志,我注意到文件系统每次机器启动时都会出现此错误:

kernel: [79137.275531] EXT4-fs (dm-2): warning: mounting fs with errors, running e2fsck is recommended

我很惭愧自己没注意 :(


然后,我尝试使用每个备份超级块(一个接一个)进行挂载,每次尝试都会在我的日志中留下以下内容:

EXT4-fs (dm-2): ext4_check_descriptors: Checksum for group 0 failed (26534!=65440)
EXT4-fs (dm-2): ext4_check_descriptors: Checksum for group 1 failed (38021!=36729)
EXT4-fs (dm-2): ext4_check_descriptors: Checksum for group 2 failed (18336!=39845)
...
EXT4-fs (dm-2): ext4_check_descriptors: Checksum for group 11911 failed (28743!=44098)
BUG: soft lockup - CPU#0 stuck for 23s! [mount:2939]


尝试重新启动 e2fsck 会导致:

# e2fsck /dev/mapper/candybox 
e2fsck 1.41.14 (22-Dec-2010)
e2fsck: Group descriptors look bad... trying backup blocks...
candy: recovering journal
e2fsck: unable to set superblock flags on candy


此时,我决定最好订购更多驱动器并使用ddrescue 现在两周后制作的图像,我在 .img 文件中有一个 luks 分区的图像。

# ls -lh
total 14T
-rw-r--r-- 1 root root 14T Oct 25 01:57 candybox.img
-rw-r--r-- 1 root root 271 Oct 20 14:32 candybox.logfile

尝试了无数次,用尽了所有能在网上找到的方法,我还是无法强制 e2fsck 对镜像做任何事情,所以我使用了,mkfs.ext4 -L candy candybox.img -m 0 -S并且能够在没有日志的情况下以只读方式挂载脏文件系统,并恢复 960G 的数据。它给出了各种错误,比如各种目录不存在等等,但我能够得到一些东西。这给了我一些希望!

然后我再次运行 e2fsck,它必须重新创建根 inode 并给出大量更正组计数列表,我接受了根 inode 创建,并对其他所有问题都说“不”,留下一个完全空的文件系统。再次运行,对所有问题都说“是”,结果相同,但现在是一个“干净”但空的文件系统。

extundelete 给了我0 recoverable inodes found.

现在我又陷入困境了,除了使用 photorec 之类的工具,我实在想不出其他方法,但这会让我陷入文件系统大小的混乱之中。

如果我能得到任何关于如何恢复更多文件的建议或想法,我愿意从原始阵列重新复制图像并重新开始。

我希望能够提供已运行命令的更详细日志,但是除了记录到 syslog 的内容之外,输出内容都被长时间滚动传递,并且由于发生这种情况的时间范围,我的记忆并不那么详细。

任何帮助是极大的赞赏!

10 月 27 日更新

我已经完全重新复制了图像,以便再次开始测试,以下是迄今为止的输出。复制过程:

[root@gamma rescue]# nbd-client 172.16.10.204 2000 /dev/nbd0
Negotiation: ..size = 14307292MB
bs=1024, sz=15002283540480 bytes
[root@gamma rescue]# cryptsetup luksOpen /dev/nbd0 candybox
Enter passphrase for /dev/nbd0: 
[root@gamma mnt]# pvcreate /dev/md5
  Physical volume "/dev/md5" successfully created
[root@gamma mnt]# pvscan
  PV /dev/md5                      lvm2 [18.19 TiB]
  Total: 1 [18.19 TiB] / in use: 0 [0   ] / in no VG: 1 [18.19 TiB]
[root@gamma mnt]# vgcreate vg-rescue /dev/md5
  Volume group "vg-rescue" successfully created
[root@gamma mnt]# lvcreate --size 15T --name lv-rescue vg-rescue
  Logical volume "lv-rescue" created
[root@gamma mnt]# mkfs.xfs /dev/vg-rescue/lv-rescue 
log stripe unit (524288 bytes) is too large (maximum is 256KiB)
log stripe unit adjusted to 32KiB
meta-data=/dev/vg-rescue/lv-rescue isize=256    agcount=33, agsize=125828992 blks
         =                       sectsz=512   attr=2
data     =                       bsize=4096   blocks=4026531840, imaxpct=5
         =                       sunit=128    swidth=640 blks
naming   =version 2              bsize=4096   ascii-ci=0
log      =internal log           bsize=4096   blocks=521728, version=2
         =                       sectsz=512   sunit=8 blks, lazy-count=1
realtime =none                   extsz=4096   blocks=0, rtextents=0
[root@gamma mnt]# mount /dev/vg-rescue/lv-rescue rescue/
[root@gamma rescue]# ddrescue /dev/mapper/candybox candybox.img candybox.ddlog


Press Ctrl-C to interrupt
Initial status (read from logfile)
rescued:         0 B,  errsize:       0 B,  errors:       0
Current status
rescued:    13194 GB,  errsize:   1807 GB,  current rate:        0 B/s
   ipos:    13194 GB,   errors:       1,    average rate:   73528 kB/s
   opos:    13194 GB,     time from last successful read:      44 s
^Clitting failed blocks... 
Interrupted by user
## Network hung, had to try again here
[regan@gamma ~]$ sudo nbd-client -d /dev/nbd0
Disconnecting: que, disconnect, Error: Ioctl failed: Invalid argument

Exiting.
[regan@gamma ~]$ sudo nbd-client 172.16.10.204 2000 /dev/nbd0
Negotiation: ..size = 14307292MB
bs=1024, sz=15002283540480 bytes

[root@gamma rescue]# ddrescue -r 2 /dev/mapper/candybox candybox.img candybox.ddlog


Press Ctrl-C to interrupt
Initial status (read from logfile)
rescued:    15002 GB,  errsize:   7426 kB,  errors:      60
Current status
rescued:    15002 GB,  errsize:       0 B,  current rate:    77529 kB/s
   ipos:    15002 GB,   errors:       0,    average rate:    69297 kB/s
   opos:    15002 GB,     time from last successful read:       0 s
Finished                       

[root@gamma rescue]# lvcreate -l 100%FREE -s -n rescue_snap /dev/vg-rescue/lv-rescue 
  Logical volume "rescue_snap" created
[root@gamma rescue]# cd ..
[root@gamma mnt]# mount -o remount,ro rescue/
[root@gamma mnt]# mkdir rescue_snap
[root@gamma mnt]# mount -o nouuid /dev/vg-rescue/rescue_snap rescue_snap
[root@gamma mnt]# cd rescue_snap/
[root@gamma rescue_snap]# ls
candybox.ddlog  candybox.img

凌乱的:

[root@gamma rescue_snap]# mkfs.ext4 -L candy candybox.img -m 0 -S
mke2fs 1.41.10 (10-Feb-2009)
candybox.img is not a block special device.
Proceed anyway? (y,n) y
Filesystem label=candy
OS type: Linux
Block size=4096 (log=2)
Fragment size=4096 (log=2)
Stride=0 blocks, Stripe width=0 blocks
915668992 inodes, 3662666368 blocks
0 blocks (0.00%) reserved for the super user
First data block=0
Maximum filesystem blocks=4294967296
111776 block groups
32768 blocks per group, 32768 fragments per group
8192 inodes per group
Superblock backups stored on blocks: 
    32768, 98304, 163840, 229376, 294912, 819200, 884736, 1605632, 2654208, 
    4096000, 7962624, 11239424, 20480000, 23887872, 71663616, 78675968, 
    102400000, 214990848, 512000000, 550731776, 644972544, 1934917632, 
    2560000000

Skipping journal creation in super-only mode
Writing superblocks and filesystem accounting information: done

This filesystem will be automatically checked every 26 mounts or
180 days, whichever comes first.  Use tune2fs -c or -i to override.

[root@gamma rescue_snap]# mount -o loop candybox.img /mnt2
[root@gamma rescue_snap]# df -h
Filesystem            Size  Used Avail Use% Mounted on
/dev/md2              147G  138G  3.1G  98% /
tmpfs                  16G  5.7M   16G   1% /dev/shm
/dev/md0              494M  199M  276M  42% /boot
/dev/sdc1             1.8T  979G  763G  57% /mnt/macmirror
/dev/sdj1             1.8T  970G  771G  56% /mnt/usbrescue
/dev/mapper/vg--rescue-lv--rescue
                       15T   14T  1.4T  91% /mnt/rescue
/dev/mapper/vg--rescue-rescue_snap
                       15T   14T  1.4T  91% /mnt/rescue_snap
/mnt/rescue_snap/candybox.img
                       14T   15M   14T   1% /mnt2

## Even though it says only 15M is used, I was able to rsync 960G to /mnt/usbrescue

[root@gamma rescue_snap]# cd /mnt2/
[root@gamma mnt2]# ls -l
ls: cannot access Fedora-19-x86_64-DVD: Input/output error
ls: cannot access rsync_batch: Input/output error
ls: cannot access shell1: Input/output error
ls: cannot access New Folder (2): Input/output error
ls: cannot access shell2: Input/output error
ls: cannot access revolution: Input/output error
ls: cannot access mail: Input/output error
ls: cannot access testing: Input/output error
ls: cannot access export: Input/output error
ls: cannot access ben_backup_20130903: Input/output error
total 160488672
drwxr-xr-x     2 regan regan        4096 Sep  3 20:16 100MEDIA
drwxr-xr-x    19 regan regan        4096 Sep 26 05:18 android
d??????????    ? ?     ?               ?            ? ben_backup_20130903
-rw-rw-r--     1 regan regan       12126 Jan  4  2013 durations.txt
d??????????    ? ?     ?               ?            ? export
drwxrwxr-x    10 regan regan        4096 Dec 29  2012 family-pc_20121229
d??????????    ? ?     ?               ?            ? Fedora-19-x86_64-DVD
-rw-r--r--     1 regan regan 72116729363 Sep 30 04:39 gamma_backup_20130928.tgz
-rw-rw-r--     1 regan regan 55606528323 Jul 27  2011 gamma_tar_20110727.tbz2
-rw-rw-r--     1 regan regan        3839 Sep 27  2012 Good Quality2.plist
-rw-rw-r--     1 regan regan        4663 Oct  7  2012 Good Quality3.plist
-rw-rw-r--     1 regan regan        3852 Sep 26  2012 Good Quality.plist
drwxr-xr-x     7 regan regan        4096 Nov 13  2012 grok
d??????????    ? ?     ?               ?            ? HardDisks
-rwxr--r--     1 regan regan       54248 Mar 16  2013 IMAG0868.jpg
-rwxr--r--     1 regan regan       51156 Mar 16  2013 IMAG0869.jpg
-rwxr--r--     1 regan regan       85912 Mar 16  2013 IMAG0870.jpg
-rwxr--r--     1 regan regan       76875 Mar 16  2013 IMAG0872.jpg
-rwxr--r--     1 regan regan       68451 Mar 16  2013 IMAG0873.jpg
-rwxr--r--     1 regan regan       59587 Mar 16  2013 IMAG0874.jpg
-rwxr--r--     1 regan regan       81232 Mar 16  2013 IMAG0875.jpg
-rwxr--r--     1 regan regan       44211 Mar 16  2013 IMAG0876.jpg
-rwxr--r--     1 regan regan       41660 Mar 16  2013 IMAG0877.jpg
-rwxr--r--     1 regan regan       36778 Mar 16  2013 IMAG0878.jpg
-rwxr--r--     1 regan regan       76964 Mar 16  2013 IMAG0879.jpg
-rwxr--r--     1 regan regan       81876 Mar 16  2013 IMAG0880.jpg
-rwxr--r--     1 regan regan     1568002 Mar 16  2013 IMAG0953.jpg
-rwxr--r--     1 regan regan     1548566 Mar 16  2013 IMAG0954.jpg
-rwxr--r--     1 regan regan     1351743 Mar 16  2013 IMAG0955.jpg
-rwxr--r--     1 regan regan     1750128 Mar 16  2013 IMAG0956.jpg
-rwxr--r--     1 regan regan     1694378 Mar 16  2013 IMAG0957.jpg
-rwxr--r--     1 regan regan     1277128 Mar 16  2013 IMAG0958.jpg
-rwxr--r--     1 regan regan     1467452 Mar 16  2013 IMAG0965.jpg
-rwxr--r--     1 regan regan     1595903 Mar 16  2013 IMAG0966.jpg
-rwxr--r--     1 regan regan     1372444 Mar 16  2013 IMAG0967.jpg
-rwxr--r--     1 regan regan     1698010 Mar 16  2013 IMAG0968.jpg
-rwxr--r--     1 regan regan     1550641 Mar 16  2013 IMAG0969.jpg
-rwxr--r--     1 regan regan     1333768 Mar 16  2013 IMAG0970.jpg
-rwxr--r--     1 regan regan     1432347 Mar 16  2013 IMAG1010.jpg
-rwxr--r--     1 regan regan     1668159 Mar 16  2013 IMAG1013.jpg
-rwxr--r--     1 regan regan     1657058 Mar 16  2013 IMAG1014.jpg
-rwxr--r--     1 regan regan     1496547 Mar 16  2013 IMAG1016.jpg
-rwxr--r--     1 regan regan     1609156 Mar 16  2013 IMAG1017.jpg
-rwxr--r--     1 regan regan     1604832 Mar 16  2013 IMAG1019.jpg
-rwxr--r--     1 regan regan     2048916 Mar 16  2013 IMAG1073.jpg
-rwxr--r--     1 regan regan     2006024 Mar 16  2013 IMAG1074.jpg
-rwxr--r--     1 regan regan     1926686 Mar 16  2013 IMAG1075.jpg
-rw-r--r--     1 regan regan     1583090 Jul 14 21:15 IMAG1565.jpg
-rw-r--r--     1 regan regan     1435031 Sep 22 05:19 IMAG1762.jpg
-rw-r--r--     1 regan regan     1531602 Sep 22 05:19 IMAG1763.jpg
-rw-r--r--     1 regan regan     1450926 Sep 22 05:19 IMAG1764.jpg
-rw-r--r--     1 regan regan     1336103 Sep 23 21:31 IMAG1765.jpg
-rw-r--r--     1 regan regan     1235885 Sep 23 21:32 IMAG1766.jpg
-rw-r--r--     1 regan regan     1224376 Sep 23 21:32 IMAG1767.jpg
-rw-r--r--     1 regan regan     1235229 Sep 23 21:32 IMAG1768.jpg
drwxrwxr-x     2 regan regan        4096 Mar  9  2013 jakeanmal
-rw-rw-rw-     1 regan regan      115228 Oct 29  2009 jj_nas
drwx------.    2 root  root        16384 Nov  8  2012 lost+found
-rw-r--r--     1 regan regan  3123877728 Nov  6  2010 luridmirror_20090806.tar.xz
-rw-r--r--     1 regan regan  2877033943 Mar  1  2013 macabre_20130301.tgz
d??????????    ? ?     ?               ?            ? mail
-rw-r--r--     1 root  root         6771 Aug 10  2009 mail_mirror
-rw-------     1 regan regan 21913047552 Apr  4  2013 mallorys_hdd.vbox.img
d??????????    ? ?     ?               ?            ? MSDN
-rw-r--r--     1 regan regan        8572 May 10  2010 Music
d??????????    ? ?     ?               ?            ? New Folder (2)
drwxrwxrwx    24 regan regan        4096 Mar 22  2013 onyx
drwxr-xr-x   231 regan regan       24576 Sep 30 09:29 ptp
-rwxr--r--     1 regan regan      483328 Jan 26  2013 putty.exe
d??????????    ? ?     ?               ?            ? revolution
-rw-r--r--     1 root  root   6272757760 Oct 16  2012 root.tar
d??????????    ? ?     ?               ?            ? rsync_batch
drwxrwxr-x     2 regan regan       12288 Oct  6 04:09 saber
drwxrwxr-x     2 regan regan      188416 Sep 25 04:20 session_tmp
d??????????    ? ?     ?               ?            ? shell1
d??????????    ? ?     ?               ?            ? shell2
d??????????    ? ?     ?               ?            ? testing
drwxrwxr-x     3 regan regan        4096 Oct  7  2012 tofix
-rwxr--r--     1 regan regan    64991966 Jan  2  2013 VIDEO0041.mp4

[root@gamma mnt2]# cd ..
[root@gamma /]# umount /mnt2
[root@gamma /]# cd /mnt/rescue_snap/
[root@gamma rescue_snap]# e2fsck candybox.img 
e2fsck 1.41.10 (10-Feb-2009)
Backing up journal inode block information.

candy contains a file system with errors, check forced.
Resize inode not valid.  Recreate<y>? yes

Pass 1: Checking inodes, blocks, and sizes
Pass 2: Checking directory structure
Pass 3: Checking directory connectivity
Root inode not allocated.  Allocate<y>? yes

/lost+found not found.  Create<y>? yes

Pass 4: Checking reference counts
Pass 5: Checking group summary information
Block bitmap differences:  -(9252--9255) <Snip a few dozen MB of text> -(3662666237--3662666238) -3662666240 -(3662666242--3662666244) -(3662666247--3662666249) -3662666253 -(3662666255--3662666256) -3662666259 -3662666262 -3662666264 -(3662666268--3662666271) -3662666276 -3662666281 -3662666285 -3662666294 -(3662666296--3662666297) -3662666301 -3662666307 -3662666309 -3662666311 -3662666313 -3662666316 -(3662666318--3662666319) -3662666324 -(3662666326--3662666328) -(3662666331--3662666332) -3662666334 -(3662666341--3662666342) -3662666344 -(3662666346--3662666347) -3662666349 -(3662666351--3662666352) -3662666354 -3662666357 -3662666362 -(3662666366--3662666367)
Fix<y>? yes

Free blocks count wrong for group #0 (23517, counted=23516).
Fix<y>? 

Free blocks count wrong (3605188902, counted=3605188901).
Fix<y>? yes

Free inodes count wrong for group #0 (8191, counted=8190).
Fix<y>? yes

Directories count wrong for group #0 (1, counted=2).
Fix<y>? yes

Free inodes count wrong (915668991, counted=915668990).
Fix<y>? yes


candy: ***** FILE SYSTEM WAS MODIFIED *****
candy: 2/915668992 files (0.0% non-contiguous), 57477467/3662666368 blocks
[root@gamma rescue_snap]# mount -o loop candybox.img /mnt2
[root@gamma rescue_snap]# ls -l /mnt2
total 4
drwx------ 2 root root 4096 Oct 27 19:33 lost+found
[root@gamma rescue_snap]# 

请注意,我现在在快照中有了备份图像,因此如果有人有想法,我可以反复尝试理论......

答案1

我首选的文件系统数据恢复工具是UFS 探索者

我不知道 LUKS 加密会对此产生怎样的影响,但请下载 UFS Explorer 评估版并尝试常规最佳实践。针对您的图像运行它,看看是否更容易恢复。该工具将为您提供目录结构的良好视图,并提供一些搜索数据的功能。

另请参阅:如何恢复出现“超级块读取失败”的 XFS 文件系统

相关内容