[SOLVED] disk space trouble

[kris-fr]

Hi all !

I am facing a strange disk space problem :

On first host i have 200GB of datas that i want to copy to a second host on an empty and freshly formated 500GB hard drive.
But during the copy process i have a "No space left on device" error. I've checked and it's right : my 500GB hard drive is 100% full....i can't understand why my 500GB is full before having copied all the 200GB of datas !??

Here is my technical architecture :
Source drive is a standard Ext3 File System
Target drive is a XFS File System
Target drive uses LVM :
sdb1 -> XFS File system -> size 200GB
sdc1 -> XFS File system -> size 200GB
sdd1 -> XFS File system -> size 100GB
On the target host i have created a Volume Group for sdb1, sdc1 and sdd1 and then i have created a Logical Volume.
I can see with "df" command that my logical volume has 500GB capacity.

Both source and target drives have 4K block size.

Note : source host is a hardware Linux box, target host is a virtual host hosted by VMWare on ESX v3.5

Any Guru help will be apreciate !

Regards

[reed9]

Please post the output of

su -c 'fdisk -l'
df -h

[kris-fr]

Result of "fdisk -l" on target host :

Disk /dev/sda: 10.7 GB, 10737418240 bytes
255 heads, 63 sectors/track, 1305 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0x0006d926

   Device Boot      Start         End      Blocks   Id  System
/dev/sda1               1          62      497983+  82  Linux swap / Solaris
/dev/sda2              63        1305     9984397+  83  Linux

Disk /dev/sdb: 214.7 GB, 214748364800 bytes
255 heads, 63 sectors/track, 26108 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0x0006030b

   Device Boot      Start         End      Blocks   Id  System
/dev/sdb1               1       26108   209712478+  83  Linux

Disk /dev/sdc: 214.7 GB, 214748364800 bytes
255 heads, 63 sectors/track, 26108 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0x0001ceb3

   Device Boot      Start         End      Blocks   Id  System
/dev/sdc1               1       26108   209712478+  83  Linux

Disk /dev/sdd: 107.3 GB, 107374182400 bytes
255 heads, 63 sectors/track, 13054 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0x000dfc3e

   Device Boot      Start         End      Blocks   Id  System
/dev/sdd1               1       13054   104856223+  83  Linux

Disk /dev/dm-0: 536.7 GB, 536766054400 bytes
255 heads, 63 sectors/track, 65258 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0x00000000

Disk /dev/dm-0 doesn't contain a valid partition table

Result of "df -h" on target host :

Filesystem            Size  Used Avail Use% Mounted on
/dev/sda2             9.4G  1.6G  7.4G  18% /
tmpfs                 1.9G     0  1.9G   0% /lib/init/rw
udev                   10M  624K  9.4M   7% /dev
tmpfs                 1.9G     0  1.9G   0% /dev/shm
/dev/mapper/vg00-lvol1
                      500G  4.2M  500G   1% /home

[reed9]

I've checked and it's right : my 500GB hard drive is 100% full

How did you check? According to df, only 4.2MB are used out of 500GB.

[kris-fr]

I've checked with the same df command.
The drive is empty just because i deleted all datas on the target to do unitary copy for testing !

[kris-fr]

Why thinking difficult when we can think easy..?
I was looking for a block size problem, logical extends problems, but it was more simple and it was not a problem.

I copy datas using recursive scp. I didn't see that some folders contain symbolic links that should not be there, but they are, for old and obsolete reasons.
As scp follows symbolic links, the "content" of symbolic links was also copied...

My datas tree is as below so i let you imagine :

/home/user1

file1
file2...
somelink -> /home/user1

/home/user2

file1
file2...
somelink -> /home/user2

...

Sorry i made a mistake !

Regards