Search results

Well, after rebooting the VM again this evening, and confirming I could interrupt the boot to get to the grub loader, now everything seems to be working fine with the console.

Tom, I'm running Windows 7, Chrome, Java 1.7.0_25. Thanks!

Just to be clear, this isn't a ctrl-s/ctrl-q issue, this persists across reboots.

I'm having a strange problem. One of my centos 6.4 KVM guests, which is otherwise working fine, doesn't accept any keyboard input on the proxmox console. My other centos 6.4 KVM guests are working fine. The tty.conf seems to be identical in all cases. Any ideas of things to try to correct this?

Switching to virtio seems to have fixed this problem, thankfully.

Well, unfortunately it died again last night, even with the limit in place for vzdump. Is there a way to specify, for a specific VM, which backup pool is used (for the automatic nightly backup)? I'm still very confused as to how an NFS problem can cause the KVM being backed up to exit. EDIT...

Worked to local storage: INFO: starting new backup job: vzdump 109 --remove 0 --mode snapshot --compress lzo --storage local --node proxmox1 INFO: Starting Backup of VM 109 (qemu) INFO: status = running INFO: backup mode: snapshot INFO: ionice priority: 7 INFO: creating archive...

Well, I spoke too soon, disabling HA on that VM did not fix it like it had fixed one other VM that had exhibited this behavior. Here is tonight's log: 109: Jun 25 05:16:55 INFO: Starting Backup of VM 109 (qemu) 109: Jun 25 05:16:55 INFO: status = running 109: Jun 25 05:16:56 INFO: backup...

One thing to add...if I disable the HA management on a VM exhibiting this problem, the problem stops occurring. Would it be helpful to disable HA on this VM for an evening to see a successful backup log? I should also add, there are HA VMs which never have this problem.

The disk is raw LVM, hosted on a DRBD-based volume.

Some nights when our backups run, one of the HA KVMs will die, and to make things worse it isn't brought up on the other node. Here's an example: 109: Jun 23 05:20:42 INFO: Starting Backup of VM 109 (qemu) 109: Jun 23 05:20:42 INFO: status = running 109: Jun 23 05:20:44 INFO: backup...

Dietmar, Is it truly not possible, or is it something that could be implemented if we paid you? If so, what sort of fees would be involved in sponsoring this development?

I'm pretty sure he's talking about the words you post and the way that people perceive them. He is not the first person to talk about the proxmox team's "attitude", in fact I was just having a conversation with a colleague the other day about proxmox and he said "Yeah, the product is great, but...

Indeed! Fortunately for me, proxmox is currently meeting all three of those needs for me. I'm just really impressed with the project and want to see it become the "industry favored" solution, must like how Wietse did with postfix a decade ago, and Timo did with dovecot several years back.

Well, in fairness, there is nothing wrong with the suggestion (I'm not the original poster, just chiming in). I'm just trying to explain what I understand of large corporations...they will pay for support (in fact often times will exclude open source solutions from lack of official support)...

It's odd, I've only been around this forum a few weeks, but it seems like the attitude of the Proxmox team could be summed up as: "If you want to suggest ways to improve Proxmox, don't bother. Do it yourself. We're not interested in your suggestions, regardless of how good they are, or how much...

I could be wrong, but I suspect it still thinks the volume is shared.

Output of pvscan and lvscan on pveabs1?

When PVEVMUser logs into the web interface, they are able to see higher level information we would prefer them not be aware of: the overall CPU and memory utilization of the servers in the cluster. We would just prefer they not be aware of any specifics beyond their KVM. Is there a way to...

Nice! I guess I underestimated you with the "be gentle" nonsense. You're clearly capable.