pvectl vzcreate bug

[calocen]

/usr/bin/ssh -t -t -n -o BatchMode=yes 194.xxx.yyy.node2 /usr/bin/pvectl vzcreate 4049 --disk 8 --ostemplate debian-4.0-standard_4.0-3_i386.tar.gz --hostname <some.fqdn> --description Debian 4.0 (standard) --nameserver <some_dns> --nameserver 194.xxx.yyy.zzz --searchdomain <my_domain> --onboot yes --userpasswd root:not_a_password --swap 512 --ipset 194.xxx.yyy.zzz --mem 512 --cpus 1
tcgetattr: Inappropriate ioctl for device
bash: -c: line 0: syntax error near unexpected token `('
bash: -c: line 0: `/usr/bin/pvectl vzcreate 4049 --disk 8 --ostemplate debian-4.0-standard_4.0-3_i386.tar.gz --hostname <some.fqdn> --description Debian 4.0 (standard) --nameserver <some_dns> --nameserver 194.xxx.yyy.zzz --searchdomain <my_domain> --onboot yes --userpasswd root:not_a_password --swap 512 --ipset 194.xxx.yyy.zzz --mem 512 --cpus 1'
Connection to 194.xxx.yyy.node2 closed.
unable to apply VM settings -


Hi.
After upgrading to last version with this command
apt-get update && apt-get --assume-yes upgrade
I get the error showed before.
Of course I try to execute in command line with the description commented out, but I only get a connection closed

Edit: That only happen when you create a new vz in other node of the cluster. In master it's works nice. vzdump, help me to keep online.

Bad upgrade !

Thanks In Advance for your help

 

[dietmar]

What is the output of:

pveversion -v

Why did you upgrade with --assume-yes ? We do not recommend that in the upgrade docu.

- Dietmar

 

[dietmar]

You are running a cluster, and you updated all nodes in the cluster?

Please send me the output of 'pveversion -v' of all nodes.

- Dietmar

 

[dietmar]

Edit: That only happen when you create a new vz in other node of the cluster. In master it's works nice.

Ok, was finally able to reproduce the bug - will release a fix tomorrow.

- Dietmar

 

[calocen]

# pveversion -v
pve-manager: 1.0-8 (pve-manager/1.0/3451)
qemu-server: 1.0-5
pve-kernel: 2.6.24-4
pve-kvm: 75-1
pve-firmware: 1
vncterm: 0.9-1
vzctl: 3.0.22-3pve3
vzdump: 1.1-1
vzprocps: 2.0.11-1dso2
vzquota: 3.0.11-1dso1

I did an automatic upgrade after an external check of availability of new packages.

 

[calocen]

Great !
Thank you for quick response.

If any other has the same problem, there's a workaround:
- First create from template in the master node
- Then migrate the new VPS to the node that you want.

 

[calocen]

Ok, was finally able to reproduce the bug - will release a fix tomorrow.

- Dietmar

"Tomorrow" is close to come?

 

[dietmar]

sorry, too many things to do - will try to provide the fix asap.

- Dietmar

 

[dietmar]

Please can you test the new package:

# wget ftp://pve.proxmox.com/debian/dists/etch/pvetest/binary-amd64/pve-manager_1.0-9_all.deb
# dpkg -i pve-manager_1.0-9_all.deb

 

[calocen]

It works !

Well done, dieter !

# pveversion -v
pve-manager: 1.0-9 (pve-manager/1.0/3461)
qemu-server: 1.0-5
pve-kernel: 2.6.24-4
pve-kvm: 75-1
pve-firmware: 1
vncterm: 0.9-1
vzctl: 3.0.22-3pve3
vzdump: 1.1-1
vzprocps: 2.0.11-1dso2
vzquota: 3.0.11-1dso1

 

[dietmar]

Ok, moved to the main repository. It is now available via apt.

- Dietmar