Comment 30 for bug 1022901

I can confirm that the "Paused" case with IO_ERROR is happening when a user (samba client) starts to copy files (70GB of pictures with 10MB/s). It crashes almost immediately!
@Serge, the Ubuntu machine is with 1GB allocated and the Windows machine is with 2GB allocated. The host is 4GB. I had a look at this possibility already, but it is not the case.
At the time of writing this, I saw that the space on host's partition is finished and now it cannot start at all, becoming Paused right after the guest tries to write something. Well, OK, but the image size by definition is this:

root@nobel-host1:/home/images# qemu-img info ubuntu-pdc-vdb.img
image: ubuntu-pdc-vdb.img
file format: qcow2
virtual size: 1.7T (1825361100800 bytes)
disk size: 1.7T
cluster_size: 65536
Snapshot list:
ID TAG VM SIZE DATE VM CLOCK
11 snapshot.sh 0 2012-07-06 00:02:02 00:00:00.000
12 snapshot.sh 0 2012-07-07 00:02:02 00:00:00.000
13 snapshot.sh 0 2012-07-10 00:02:02 00:00:00.000
14 snapshot.sh 0 2012-07-12 00:06:01 00:00:00.000
15 snapshot.sh 0 2012-07-13 00:06:03 00:00:00.000

But the real size is (as I am observing from a while) increasing. I don't know if it is from snapshot taking into the image file itself? Also now it didn't do any snapshot, the only thing is copying files into the guest. Isn't the size of the image supposed to be fixed at that, mentioned above? Why is it now this:

root@nobel-host1:/home/images# du -sm ubuntu-pdc-vdb.img
1785438 ubuntu-pdc-vdb.img

and at the beginning it was this:
1741809 ubuntu-pdc-vdb.img

It has grown with over 40GB for about half a month. I'm doing a snapshot every evening at working days, but am deleting the old ones too, leaving only 5 snapshots at a time.