11

We have SUSE Linux OS:

$ uname -a Linux 3.0.13-0.27-default #1 SMP Wed Feb 15 13:33:49 UTC 2012 x86_64 x86_64 x86_64 GNU/Linux 

Only for a specific file the file creation fails with error no space left on device. We are able to create other files, what has gone wrong? In the current scenario:

  • Inode usage is only 47%

    $ df -i . Filesystem Inodes IUsed IFree IUse% Mounted on /dev/mapper/appvg-dsfs 39321600 18218000 21103600 47% /dsfs 
  • Disk usage is only 82%

    $ df -h . Filesystem Size Used Avail Use% Mounted on /dev/mapper/appvg-dsfs 591G 457G 105G 82% /dsfs 
  • File creation of one specific file fails in this case JDE_GL_INCR_STG_201504_181.log.load.058

    $ echo "a" > JDE_GL_INCR_STG_201504_181.log.load.058 -bash: JDE_GL_INCR_STG_201504_181.log.load.058: No space left on device echo "a" > testfile $ mv JDE_GL_INCR_STG_201504_181.log.load.058 tempfile mv: cannot stat `JDE_GL_INCR_STG_201504_181.log.load.058': No such file or directory 
  • File creation of other files is possible

12
  • 4
    Is any SELinux profile enabled? Commented May 6, 2015 at 13:20
  • 3
    Not sure if this matters, but what filesystem are you using? For example XFS has some issues with full partitions, although your disk is not that full. Commented May 6, 2015 at 13:27
  • What file system is it? (Maybe it is a hardware error? What does dmesg say?) Commented May 6, 2015 at 13:29
  • 3
    Could this be a hashing issue as discussed at serverfault.com/questions/539536/… ? Commented May 6, 2015 at 13:54
  • 5
    I'd fsck.ext3 it. Commented May 6, 2015 at 16:04

4 Answers 4

1

really do look like hash-collision issue. Have u tried creating this exact same file in any other empty directory on the same partition to test if it is really hash-collision issue or something else !!

1

I have seen issues of this nature when I have mounted a faulty or corrupted drive. If you have smartctl on the host, run it against the device (smartctl -a /dev/<device>). Different drives have different data, but look for things like: uncorrected errors, multi_zone_error_rate, current_pending_sector, raw_read_error_rate, or if it says the drive has failed. Those are all indicators that the is not in the best state.

1

Maybe you run out of Semaphores. Try this as root

ipcs -s 

and check if there are a lot of semaphores owned by the same user. If so run the following command

ipcs -s |grep username |awk '{print $2}' 

in the above command replace the username withn the actual user.

-2

Firstly check the size of the file you are trying to create. The file might be very big. Ok I know the chances of the above being the issue may be low, rename the file to a different name or create it with different file name..

1
  • 1
    Please read again my question. The file creation command is mentioned there. echo "a" > JDE_GL_INCR_STG_201504_181.log.load.058 By this method the file size is not over a kb. Even touch gives same error touch JDE_GL_INCR_STG_201504_181.log.load.058 Commented May 7, 2015 at 6:23

You must log in to answer this question.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.