9

I am using a VM on Google Compute Cloud. I grew my disk from 10G to 200G.

I followed the exact steps here: https://cloud.google.com/compute/docs/disks#repartitionrootpd

To summarize:

  • I ran fdisk, removed the only partition, created a new one of the full size, same start / new end, same device ID
  • I rebooted the instance
  • I resized my filesystem using sudo xfs_growfs / (I am running CentOS 7)

After this I untar a 3.5G archive in a /opt subdirectory which, after a few minutes, ended with :

Cannot mkdir: No space left on device

I can check that the space is here and it seems (to me at least) that it should be available everywhere

# df -h
Filesystem      Size  Used Avail Use% Mounted on
/dev/sda1       200G   13G  188G   7% /
devtmpfs        1.9G     0  1.9G   0% /dev
tmpfs           1.9G     0  1.9G   0% /dev/shm
tmpfs           1.9G  8.3M  1.8G   1% /run
tmpfs           1.9G     0  1.9G   0% /sys/fs/cgroup

Now, with this exact configuration, a simple cp command on a 50Mb directory also returns:

cp: cannot create regular file ‘toto/conf/server.xml’: No space left on device

I had many small files in my tar so I thought about a inode limitation, but:

# df -ih
Filesystem     Inodes IUsed IFree IUse% Mounted on
/dev/sda1        200M  100K  200M    1% /
devtmpfs         462K   285  462K    1% /dev
tmpfs            463K     1  463K    1% /dev/shm
tmpfs            463K   309  463K    1% /run
tmpfs            463K    13  463K    1% /sys/fs/cgroup

It is as if my new disk space is not available. Because I have the feeling that it approximately stopped at my former 10G disk limitation.

I have no idea what to do now.

1

2 Answers 2

12

I managed to make it work through the command:

mount -o remount,inode64 /

Apparently this is a regression in CentOS kernel 3.7 to 3.17 and I am in 3.10.

Here is the relevant link: http://xfs.org/index.php/XFS_FAQ#Q:_Why_do_I_receive_No_space_left_on_device_after_xfs_growfs.3F

1
  • 1
    You can accept your own answer as well and this is what I suggest to you (to not let open this question until the eternity).
    – peterh
    Nov 29, 2014 at 0:19
1

Change the mount options to use inode64.

This issue seems like the one from the XFS FAQ

http://xfs.org/index.php/XFS_FAQ#Q:_Why_do_I_receive_No_space_left_on_device_after_xfs_growfs.3F

2
  • Answered in comment earlier because I wasn't sure If this was the right answer since inode64 is the default for XFS in the kernel included in CentOS 7
    – austinian
    Nov 29, 2014 at 0:26
  • I think this is "Google compute cloud" related as it is the Google infrastructure which connects the disk to my VM. I did not use anything else but the defaults Nov 29, 2014 at 6:07

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .