Help assessing the output of fsck on my external drive

geo909 · 2014-10-01 04:43:17

EDIT: It's Terabytes, not Gigabytes

Hi all,

I have a ~~3GB~~ 3 TB external hard drive in ext4. It's new but lately it has often been giving me
input/output errors. That said, there was a heavy usage at the time, with two external
disks being written with data. Also, I had once unplugged the drive by mistake while data
was being written on it.

Could somebody tell me what conclusions can we have from the following result of fsck?
I'm curious if the disk is damaged or something.

Thanks in advance.

jorge@flamingo:~$ sudo fsck.ext4 -cDfty -C 0 /dev/sdb1
e2fsck 1.42.12 (29-Aug-2014)
BRICK: recovering journal

Checking for bad blocks (read-only test):   0.43% done, 7:22   0.43% done, 7:24 elapsed. (0/0/0 errors)   <---- That went on for about a day

BRICK: Updating bad block inode.
Pass 1: Checking inodes, blocks, and sizes
Pass 2: Checking directory structure                                           
Pass 3: Checking directory connectivity                                        
/lost+found not found.  Create? yes

Pass 3A: Optimizing directories
Pass 4: Checking reference counts                                              
Pass 5: Checking group summary information                                     
Free blocks count wrong (667461777, counted=667500230).                        
Fix? yes

Free inodes count wrong (183133100, counted=183133113).                        
Fix? yes


BRICK: ***** FILE SYSTEM WAS MODIFIED *****
BRICK: 15431/183148544 files (15.4% non-contiguous), 65066042/732566272 blocks
Memory used: 608k/704k (470k/139k), time: 90243.44/14.23/ 0.19
I/O read: 35MB, write: 21MB, rate: 0.00MB/s

Last edited by geo909 (2014-10-01 20:32:43)

graysky · 2014-10-01 09:01:26

Some descripencies were found and fixed. You should run fsck in bad blocks mode to see if you have physical damage.

e2fsck -vcck /dev/sdXY

Look for reallocated sectors with smartmontools as well...

edit: 3G? Must be a USB. Dunno how bad sectors are treated with those...

Last edited by graysky (2014-10-01 09:02:50)

geo909 · 2014-10-01 20:29:33

graysky wrote:

Some descripencies were found and fixed. You should run fsck in bad blocks
mode to see if you have physical damage.
e2fsck -vcck /dev/sdXY
Look for reallocated sectors with smartmontools as well...
edit: 3G? Must be a USB. Dunno how bad sectors are treated with those...

Yes, it's an external USB drive in ext4. But I thought the bad sectors were irrelevant to
the disk being an external USB or internal?

In any case, I'll run e2fsck as you told me and will report back.

Thanks.

Last edited by geo909 (2014-10-01 20:30:22)

graysky · 2014-10-01 20:30:21

Is it solid state or rotating? Maybe it doesn't matter and I am silly.

geo909 · 2014-10-01 20:33:26

graysky wrote:

Is it solid state or rotating? Maybe it doesn't matter and I am silly.

Oops! I'm sorry, all this time I meant 3 TB, not gigabytes.. :S Just realized it.
It's a rotating 3 TB disk.

Last edited by geo909 (2014-10-01 20:33:46)

graysky · 2014-10-01 20:51:11

OK... the command I posted will do a scan of the entire disk... for 3TB, that will likely take days. Do you see any reallocated sectors from `smartctl --all /dev/sdX | grep -i reall` ? If you do, scanning the disk might be a waste of your time. Once you get them, more tend to follow. Probably better to RMA the drive.

geo909 · 2014-10-02 02:33:04

graysky wrote:

OK... the command I posted will do a scan of the entire disk... for 3TB, that will likely take days. Do you see any reallocated sectors from `smartctl --all /dev/sdX | grep -i reall` ? If you do, scanning the disk might be a waste of your time. Once you get them, more tend to follow. Probably better to RMA the drive.

smartctl says the health status is OK. So I'll start running e2fsck.

On another note, after googling around, I started suspecting the power too. I live in an old house, with very few outlets,
so I have one outlet for: a monitor; speakers; a laptop; a printer/scanner; a powered usb hub connected to an external
disk and an mp3 player; a desk lamp, and this hard disk.

Also, the last time the disk gave me a I/O error was when I was doing a very large synchronization with unison in two disks
simultaneously. With normal use I haven't had anything unusual so far, so maybe it's just that I've been asking too much
from my drives.

Last edited by geo909 (2014-10-02 02:35:10)

geo909 · 2014-10-02 03:08:08

I found the culprit.. The stupid enclosure has an automatic sleep function. If not
accessed, it will sleep in 5 minutes and will give an I/O error if I try to access it
afterwards. This model (ORICO 7618US3) does not come with any win software
that could turn the darn function off.

I googled around and the only relevant thing I found is the sdparm tool. I tried it
as follows:

jorge@flamingo:~$ sudo sdparm --clear STANDBY -6 /dev/sdc
    /dev/sdc: ORICO     7618US3  0100
change_mode_page: failed fetching page: Power condition

So it seems I have no luck with that..

Do you think I have any options other than maybe having a cron job or something
accessing it every 5 minutes?

Last edited by geo909 (2014-10-02 03:11:41)

Arch Linux

#1 2014-10-01 04:43:17

Help assessing the output of fsck on my external drive

#2 2014-10-01 09:01:26

Re: Help assessing the output of fsck on my external drive

#3 2014-10-01 20:29:33

Re: Help assessing the output of fsck on my external drive

#4 2014-10-01 20:30:21

Re: Help assessing the output of fsck on my external drive

#5 2014-10-01 20:33:26

Re: Help assessing the output of fsck on my external drive

#6 2014-10-01 20:51:11

Re: Help assessing the output of fsck on my external drive

#7 2014-10-02 02:33:04

Re: Help assessing the output of fsck on my external drive

#8 2014-10-02 03:08:08

Re: Help assessing the output of fsck on my external drive

Board footer