Arch Linux

harlekin

Member

From: Germany

Registered: 2006-07-13

Posts: 408

sshfs broke my filesystem (?)

Hello.
Yesterday I had to unpack a archive file located at my laptop to my desktop pc. I had to unpack it because neither on my laptop nor on the desktop pc was enough space to hold both the archive and the unpacked version of this file.

I tried to accomplish the task with ssh without any success and I stumbled over sshfs which seems great for my particular purpose.

$ sshfs desktop:/mnt/inc ./foo

This command mounts the partition /mnt/inc located at my desktop to foo on my laptop. It seems to work great. I just went on unpacking my rar file:

unrar x file.rar ./foo/

No problem at all. This morning, I wanted to play some mp3s located at my desktop pc and all but one failed (ncmpc doesn't give any specific error). So after  listing the directories in /mnt/inc failed too, I took a look at the log files.

dmesg | grep hdb contains: (hdb is the broken hdd)

hdb: drive is not ready for command
hdb: status error: status=0x59 { DriveReady SeekComplete Data Request Error }
hdb: status error: error=0x40 { UncorrectableError }, LBAsect=56098895, high=3, low=5767247, sector=243
end_request: I/O error, dev hdb, sector 243

Plenty of lines like this follow. Seems as if it is no sector readable anymore.

I found this lines in kernel.log:

EXT3-FS on hdb1, internal journal
EXT3-fs error (device: hdb1): ext3_free_blocks: Freeing blocks in system zones - Block: 19529906, count = 1
EXT3-fs error (device: hdb1), ext3_new_block: Allocating block in system zone - blocks from 3080349, length = 1
...
hdb: dma_intr: status=0x59 { DriveReady SeekComplete DataRequest Error }
hdb: dma_intr: error=0x40 { UncorrectableError }, LBAsect=68729423, high=4, low=1620559, sector=68729311
end_request: I/O error, dev hdb, sector 68729311

As well as in dmesg there are many more lines following the same scheme.

I can neither run fsck nor mount this partition any more:

# fsck.ext /dev/hdb
e2fsck 1.39 (29-May-2006)
fsck.ext3: Attempt to read block from filesystem resulted in short read while trying to open /dev/hdb
Could this be a zero-length partition?

# mount -o ro /dev/hdb1 /mnt/inc
mount: You have to specify a filesystem

# mount -o ro -t ext3 /dev/hdb1 /mnt/inc
mount: wrong fs type, bad option, bad superblock on /dev/hdb1,
       missing codepage or other error
       In some cases useful info is found in syslog - try
       dmesg | tail  or so

And dmesg gives:

hdb: status error: status=0x59 { DriveReady SeekComplete DataRequest Error }
hdb: status error: error=0x40 { UncorrectableError }, LBAsect=56098895, high=3, low=5767247, sector=0
ide: failed opcode was: 0xea
hdb: drive not ready for command
hdb: wcache flush failed!

Thankful for any suggestion on how to solve this (if this can be done) and on what caused this problem.

harlekin

Last edited by harlekin (2007-06-10 08:45:26)

---

Hail to the thief!

ataraxia

Member

From: Pittsburgh

Registered: 2007-05-06

Posts: 1,553

Re: sshfs broke my filesystem (?)

This looks like a disk failure. Probably nothing to do with sshfs, just weird timing.

harlekin

Member

From: Germany

Registered: 2006-07-13

Posts: 408

Re: sshfs broke my filesystem (?)

Thanks.

Any posibility to check wether the disk is physically working? Are there any tools you can recommend?

Last edited by harlekin (2007-06-11 04:59:10)

---

Hail to the thief!

byte

Member

From: Düsseldorf (DE)

Registered: 2006-05-01

Posts: 2,046

Re: sshfs broke my filesystem (?)

UltimateBootCd has diagnosis tools from every hard disk vendor. Knoppix or others at least have smartmontools.

---

1000