Restore doesn't go all the way?

  • Thread starter Thread starter jonathanb
  • Start date Start date
J

jonathanb

Guest
Hello everyone,

I have problem after problem and now that I can mount the NFS and now how to restore it I have another problem. I used the following command to restore my windows 2008 r2 x64.

Code: 
xenhead:/mnt/nfs-mountpoint# qmrestore vzdump-qemu-101-2011_04_15-15_00_02.tar 500
Now it doesn't go all the way I think. After two tries I have a disk with three vm disks of 80GB which I can't find anywhere in the webGUI.

This is what my shell is giving:

Code: 
xenhead:/mnt/nfs-mountpoint# dir
images
vzdump-qemu-101-2011_04_15-15_00_02.log
vzdump-qemu-101-2011_04_15-15_00_02.tar
xenhead:/mnt/nfs-mountpoint# qmrestore vzdump-qemu-101-2011_04_15-15_00_02.tar 500
INFO: restore QemuServer backup 'vzdump-qemu-101-2011_04_15-15_00_02.tar' using ID 500
INFO: extracting 'qemu-server.conf' from archive
INFO: extracting 'vm-disk-ide0.raw' from archive
INFO:   Rounding up size to full physical extent 80.00 GB
INFO:   Logical volume "vm-500-disk-1" created
INFO: new volume ID is 'HDD1:vm-500-disk-1'
INFO: restore data to '/dev/hdd_sda1/vm-500-disk-1' (85899345920 bytes)
^CINFO: starting cleanup
  Can't remove open logical volume "vm-500-disk-1"
ERROR: unable to cleanup 'HDD1:vm-500-disk-1' - command '/sbin/lvremove -f hdd_sda1/vm-500-disk-1' failed with exit code 5
ERROR: restore QemuServer backup 'vzdump-qemu-101-2011_04_15-15_00_02.tar' failed - interrupted by signal
xenhead:/mnt/nfs-mountpoint# qmrestore vzdump-qemu-101-2011_04_15-15_00_02.tar 500
INFO: restore QemuServer backup 'vzdump-qemu-101-2011_04_15-15_00_02.tar' using ID 500
INFO: extracting 'qemu-server.conf' from archive
INFO: extracting 'vm-disk-ide0.raw' from archive
INFO:   Rounding up size to full physical extent 80.00 GB
INFO:   Logical volume "vm-500-disk-2" created
INFO: new volume ID is 'HDD1:vm-500-disk-2'
INFO: restore data to '/dev/hdd_sda1/vm-500-disk-2' (85899345920 bytes)
ping
^CINFO: starting cleanup
  Can't remove open logical volume "vm-500-disk-2"
ERROR: unable to cleanup 'HDD1:vm-500-disk-2' - command '/sbin/lvremove -f hdd_sda1/vm-500-disk-2' failed with exit code 5
ERROR: restore QemuServer backup 'vzdump-qemu-101-2011_04_15-15_00_02.tar' failed - interrupted by signal
xenhead:/mnt/nfs-mountpoint#
I pressed CTRL+C after an hour. A newly installed Win08R2 x64 with 80GB space isn't that hard to restore or am I wrong?

What am I doing wrong? :confused:
 
Hello Dietmar,

Thank you for your reply and my English isn't great as you already see. I meant difficult instead of hard. The harddisk that I have can read 100 MB/s and I don't know how fast the writing is. It takes more than 1.5 hours to complete this and that means that it doesn't work or that my harddisk is a bit slow. Today i'll let it run for four hours and if it isn't done after four hours then there should be something wrong with it.

The command that I'm using is correct right?
 
Yesterday it ran for more than four hours and it didn't finish. I don't have a clue why it doesn't finish. :(

When I did a search on google I saw many post about how slow the restoration was, but it doesn't take a whole day or does it?
 
The harddisk is 1.5TB and I'm using just 400GB of it. There are failed restores on the LVM which I can't delete,but it has still 1100GB left so it shouldn't be a problem. :confused:
 
So I did it again and know I tailed the syslog and it only gave me this:

Jun 2 16:54:59 xenhead vzdump[8145]: INFO: restore QemuServer backup 'vzdump-qemu-101-2011_06_01-15_00_02.tar' using ID 404

and a lot of:

Jun 2 16:55:31 xenhead pvedaemon[7791]: WARNING: Cannot encode 'meminfo' element as 'hash'. Will be encoded as 'map' instead

which was already happening before I tried to restore the backup.

as I already said I'm no Linux expert, but thats not looking good.

1.jpg2.jpg

In the screenshots above you can see the files that are in my harddisk which I want to restore. There are 6 disks where 5 of them are restores of the first (vm-101-disk-1) disk. I went to the location of these files and did the rm /herethefiles which deleted these disks, but I still see them in the webgui and it's eating my space. How can I delete these failed disks and do you see anything wrong with how I'm trying to restore my Win08R2 server disk?

I'm really feeling bad because I can't get it to work.
 
I don't know what happend, but when I tried it again I had a bsod and it suddenly worked when I came back! I'm trying to do it again now and hopefully it will work again.

I still haven't fix the problem where proxmox still see the other VM's. How can I delete those? There are two screenshots in my previous post.
 
Strange enough the virtual machines doesn't come up in the web interface. I can't use the lvremove because I would remove the whole thing with everything in it and I still have some good working machines in it. Is there a config file where I can delete the entries from or something like that?

Thank you for your reply!
 
jonathanb said:
I can't use the lvremove because I would remove the whole thing with everything in it and I still have some good working machines in it.
Click to expand...

What do you mean by 'whole thing'? lvremove just removes a single logical volume.

jonathanb said:
Is there a config file where I can delete the entries from or something like that?
Click to expand...

I doubt there is a VM config file (else it would show up in the web interface). Please check at /etc/qemu-server/<VMID>.conf
 
On the volume I have multiple VM's that are working, but there are a couple that don't work. If I use the lvremove won't I remove all the machines with it?
 
This is what I want.

storage.png browse.png problem.png

This is what I want, but if I use the lvremove I will delete the lvm and all the VM's in it. I still need the 101VM so I can't use this method of deleting the VM's. It's also strange that I can't find these VM's in the VM manager.

If you don't understand me then please say so.


---
Jonathan
 
jonathanb said:
This is what I want, but if I use the lvremove I will delete the lvm and all the VM's in it.
Click to expand...

I do not really understand what you means. You can simple delete a single logical volume, for example:

# lvremove HDD1/vm201-disk-1

What exactly is the problem?
 
My bad. I thought I would remove every VM in there, but know I understand I only delete the VM when I use that line. I'm now figuring out how it works. Thank you for the help!
 
You must log in or register to reply here.
Top Bottom