Ls cannot open directory input output error

My system raised the I/O error when I tried to use ‘ls’ on a mounted hard disk.

I am using

hadoop@hbase1:/hddata$ uname -a
Linux hbase1 3.8.0-29-generic #42~precise1-Ubuntu SMP Wed Aug 14 16:19:23 UTC 2013 x86_64    x86_64 x86_64 GNU/Linux

hadoop@hbase1:/hddata$ df
Filesystem                      1K-blocks       Used Available Use% Mounted on
/dev/mapper/hbase2--vg-root     468028968    2715496 441532304   1% /
udev                              6081916          4   6081912   1% /dev
tmpfs                             2436652        336   2436316   1% /run
none                                 5120          0      5120   0% /run/lock
none                              6091620          0   6091620   0% /run/shm
/dev/sda1                      1922727280  867279740 957771940  48% /hddata
/dev/sdb1                          233191      27854    192896  13% /boot
10.18.103.101:/data/marketdata 1883265024 1644255232 143344640  92% /srv/data/marketdatah

The last several lines of dmesg

hadoop@hbase1:/hddata$ dmesg | tail
[316263.280056] EXT4-fs (sda1): previous I/O error to superblock detected
[316263.281326] sd 0:0:0:0: [sda] Unhandled error code
[316263.281329] sd 0:0:0:0: [sda]
[316263.281330] Result: hostbyte=DID_BAD_TARGET driverbyte=DRIVER_OK
[316263.281332] sd 0:0:0:0: [sda] CDB:
[316263.281334] Write(10): 2a 00 00 00 00 3f 00 00 08 00
[316263.281342] end_request: I/O error, dev sda, sector 63
[316263.282584] Buffer I/O error on device sda1, logical block 0
[316263.283799] lost page write due to I/O error on sda1
[316263.283842] EXT4-fs error (device sda1): ext4_find_entry:1270: inode #2: comm bash: reading directory lblock 0

Answers

• I may have found the problem.  I initially created the share by right clicking a folder name and using the NFS Sharing tab.   I just deleted that share/folder and set up a new share using the Wizard in Server Manager.   I’m able to read
  and write data to the share from my RHEL7 machine.   

  So no, this wasn’t a hardware issue with Linux.. just a config issue with 2012 which I suspected.  

  • Marked as answer by
    Mary Dong
    Tuesday, March 15, 2016 6:31 AM

Hi

I have a directory «foo» on a NTFS partition of my hard drive, which i created a couple of days ago.
After creation it worked fine and i put some files in it, but the day after i get the following error:

me@mymachine:/media/disk1$ ls -l foo
ls: cannot access foo: Input/output error

me@mymachine:/media/disk1$ ls -l
ls: cannot access foo: Input/output error
total 101
d????????? ? ? ? ? ? foo
(+ a succesful listing of other files and directories)

I get the Input/output error no matter what i try to do to the directory (rm, mv, cd, etc.), and no matter which privileges i have.

I read some threads suggesting that it is a hardware problem. I tried running ntfsfix on the partition:

me@mymachine:~$ sudo ntfsfix /dev/sda3
Mounting volume… OK
Processing of $MFT and $MFTMirr completed successfully.
NTFS volume version is 3.1.
NTFS partition /dev/sda3 was processed successfully.

I have also run the short and the long SMART test from the disk utility, resulting in the overall assessment: Disk is healthy.

It would be nice to know how to recover the data in the directory, but i have backup so that part is not crucial. I could also reformat the whole drive to ext4, which I eventually would have done in any case. Or i could just ignore the defective folder which does not harm anyone.
However if this problem is the first warning of a failing drive, I would rather get a new drive now, before some more important data suddenly disappear.

My question is now, what course of action i should take.
Is there any other diagnostics that i could try? Is it a soft- or hardware related issue. Should I begin looking for a new hard drive?

That most definitely seems like a bad hard drive.

Of course a hardware problem elsewhere could also cause data/file system corruption. Do you have reason to suspect it could be anything else? Upgraded your CPU or RAM lately?

Thanks for the reply.
Nope, no hardware changes. I’ve been running Hardy to Karmic on unchanged hardware without problems of this kind before.
I guess I should look into changing the disk. However it would be nice to be able to verify that the hardware is to blame.

Originally Posted by tjoff

Thanks for the reply.
Nope, no hardware changes. I’ve been running Hardy to Karmic on unchanged hardware without problems of this kind before.
I guess I should look into changing the disk. However it would be nice to be able to verify that the hardware is to blame.

You answered your own question. Try a Karmic.live CD and see if the problem still occurs.

Thank you for taking the time to answer me.

Regarding booting from a live-CD:

I tried it, the problem persists.
I still wonder however, what I have proved by booting from the live CD.
How should I be able to distinguish between a corrupted filesystem (in which case the problem can be solved by reformatting the drive), and a physically failing hard drive (in which case the solution is to exchange the drive)?

I was hoping to be able to make some kind of diagnostics for the hard drive, which could give me a more verbose error message than «Input/output error», and indicate the what causes the error.
I did all the tests from the disk utility tool, and none indicate any problems.

A definite proof or disproof that the hardware is failing would, apart from assuring myself that it is necessary to change the drive, come in handy as there is still a valid guarantee on the computer.

I tried to boot with windows xp, and chkdsk or whatever it is called found the errors and deleted the corrupted directory.
In ubuntu i no longer have any indication of the problem. I guess it was a software problem after all, and that ntfsfix was not able to fix the ntfs.

huhuhu,
me also encountered the same thing as yours, tjoff.

the case in me is that,
my ftp mounted directory seems can not accept
some files after cp operation… Geeez…