problem creating container on remote node

byteslo

New Member
Feb 14, 2012
4
0
1
We just created a cluster using the documentation. The cluster is set up over two different server companies in different lcoations, so the communication is over public ip, adding the cluster seemed to go fine.

CID----IPADDRESS----ROLE-STATE--------UPTIME---LOAD----MEM---DISK
1 : 64.x.x.138 M A 6 days 03:50 0.04 8% 1%
2 : 74.x.x.146 N S 01:51 0.00 6% 1%

The cluster showed sync and everything was good, and we just tried to create a container on the remote node. It is sitting at the command and not completing for about 45 mins. We tried this twice. It shouldnt take that long to create a new container correct? Any thoughts?

It does show the command on both the master and node:

master: /usr/bin/ssh -t -t -n -o BatchMode=yes 74.x.x.146 /usr/bin/pvectl vzcreate 105 --disk 8 --ostemplate local:vztmpl/ce
ntos-5-standard_5.6-1_i386.tar.gz --rootpasswd '$1$DttDOJ/J$qVQmGFHbs4g17EVngDfOrl4.' --hostname my.local.com --nameserver 4.2.2.2 --searchdomain local.com
--onboot yes --swap 1024 --mem 1024 --cpus 1 --netif 'ifname=eth0,mac=E6:32:00:7C:58:EF'

and on node: /usr/bin/perl -w /usr/bin/pvectl vzcreate 105 --disk 8 --ostemplate local:vztmpl/centos-5-standard_5.6-1_i386.tar.gz
--rootpasswd $1$DttDOJ/J$qVQmGFEV117EVngDfOrl4. --hostname my.local.com --nameserver 4.2.2.2 --searchdomain local.com --onboot yes --swap 1024 --mem 1
024 --cpus 1 --netif ifname=eth0,mac=E6:32:00:7C:58:EF


Any idea why this wouldnt be working? Please advise, thank you.
 
I am still digging around and found this post http://forum.proxmox.com/threads/7221-Proxmox-cluster-on-VPN saying that you cant run a cluster node on a different physical network, is that true? At the end of the post the last comment is that the user does in fact have it working using public ips, which is how we have it set up since security of traffic and login page is not a huge concern. I figured I would reboot the slave node and it came up fine, but then went into no-sync but the log on the master node shows it synced and is still there. I am going to give it some hours to see if it goes back to sync, and if anyone replies :) ..... thanks guys.


Feb 17 18:27:03 pvemirrorstarting cluster syncronization
Feb 17 18:27:03 pvemirrorsyncing templates
Feb 17 18:27:03 pvemirrorcluster syncronization finished (0.16 seconds (files 0.00, config 0.00))
Feb 17 18:28:03 pvemirrorstarting cluster syncronization
Feb 17 18:28:03 pvemirrorsyncing templates
Feb 17 18:28:03 pvemirrorcluster syncronization finished (0.15 seconds (files 0.00, config 0.00))
Feb 17 18:29:03 pvemirrorstarting cluster syncronization
Feb 17 18:29:03 pvemirrorsyncing templates
Feb 17 18:29:03 pvemirrorcluster syncronization finished (0.15 seconds (files 0.00, config 0.00))
Feb 17 18:30:03 pvemirrorstarting cluster syncronization
Feb 17 18:30:04 pvemirrorsyncing templates
Feb 17 18:30:04 pvemirrorcluster syncronization finished (0.19 seconds (files 0.00, config 0.00))

[TD="align: right"]862271 [/TD]

[TD="align: right"]862271 [/TD]

[TD="align: right"]862271 [/TD]

[TD="align: right"]862271 [/TD]

[TD="align: right"]862271 [/TD]

[TD="align: right"]862271 [/TD]

[TD="align: right"]862271 [/TD]

[TD="align: right"]862271 [/TD]

[TD="align: right"]862271 [/TD]

[TD="align: right"]862271 [/TD]

[TD="align: right"]862271 [/TD]

[TD="align: right"]862271 [/TD]
 
Thank you. I am not sure why that would seem to me like a downgrade no? Why would the feature be taken away to be able to use it over different geo locations? Anyhow thank you for the reply. I noticed the images directory was huge and synching, i ran a manual rsync and after that let the clustersync finish, was then able to successfully create the container on the other node. Thanks for the repy.