Hello again!
At the weekend wanted to restore a backup. It is a VM in qcow2 with 2.4TB content and 8.9TB virtual disk storage.
A restore failed after a few minutes. Fortunately, I was able to manually restore the VM as RAW.
The virtual size of the qcow2 file is approximately 8.3TB before the error appears. I assume qmrestore is not waiting long enough for the qcow2 container to be completed.
Update: qemu-img executed manually works:
At the weekend wanted to restore a backup. It is a VM in qcow2 with 2.4TB content and 8.9TB virtual disk storage.
A restore failed after a few minutes. Fortunately, I was able to manually restore the VM as RAW.
The virtual size of the qcow2 file is approximately 8.3TB before the error appears. I assume qmrestore is not waiting long enough for the qcow2 container to be completed.
Code:
()
restore vma archive: zstd -q -d -c /var/lib/vz/backups/dump/vzdump-qemu-104-2020_06_20-07_10_11.vma.zst | vma extract -v -r /var/tmp/vzdumptmp8963.fifo - /var/tmp/vzdumptmp8963
CFG: size: 383 name: qemu-server.conf
DEV: dev_id=1 size: 9663676416000 devname: drive-virtio0
CTIME: Sat Jun 20 07:10:12 2020
no lock found trying to remove 'create' lock
TASK ERROR: command 'set -o pipefail && zstd -q -d -c /var/lib/vz/backups/dump/vzdump-qemu-104-2020_06_20-07_10_11.vma.zst | vma extract -v -r /var/tmp/vzdumptmp8963.fifo - /var/tmp/vzdumptmp8963' failed: command '/usr/bin/qemu-img create -o 'preallocation=metadata' -f qcow2 /var/lib/vz/images/101/vm-101-disk-0.qcow2 9437184000K' failed: got timeoutUpdate: qemu-img executed manually works:
Bash:
# time /usr/bin/qemu-img create -o 'preallocation=metadata' -f qcow2 /var/lib/vz/images/101/vm-101-disk-0.qcow2 9437184000K
Formatting '/var/lib/vz/images/101/vm-101-disk-0.qcow2', fmt=qcow2 size=9663676416000 cluster_size=65536 preallocation=metadata lazy_refcounts=off refcount_bits=16
real 9m1.503s
user 0m17.885s
sys 0m15.180s
# ls -lh
-rw-r--r-- 1 root root 8.8T Jun 23 09:32 vm-101-disk-0.qcow2
Last edited:
