On this VM is installed pFsense (it is our virtual router/firewall). This VM has no guest agent (i can see the "No Guest Agent configured" message on the Summary tab of the VM in the Proxmox GUI). If we take another, our VM102, it is more small one (2GiB memory, 2 CPU, 4Go of hard disk), and the fail is more rare (around 5% i would say). So it is not a specific problem to one VM. We got another VM like this one (big one with docker, guest agent installed) and it can fails the same around 30%. This VM has guest agent intalled (and i can see the IP on the Summary tab of the VM in the Proxmox GUI) and can fail randomly around 30% of time i think. INFO: start failed: : Unit 100.scope already exists.Ĭommand 'qm start 100 -skiplock' failed: exit code 255įor instance our VM100 is a big one of 40GiB memory, 16 CPU, 200Go of hard disk (this VM contains a docker host and its docker containers). INFO: include disk 'virtio0' 'local-zfs:vm-100-disk-0' 900G INFO: starting new backup job: vzdump 100 -storage local -mode stop -mailto -remove 0 -node pve01 -compress lzo
I get failures on both Manual and scheduled backup when using Stop. I have just installed latest release of Proxmox on a Dell R710, with 48 GB ram, 6 x 4TB Seagate HDD under ZFS raid 2. INFO: vm is online again after 26 secondsĮRROR: Backup of VM 107 failed - start failed: : Unit 107.scope already exists. INFO: starting new backup job: vzdump 107 -storage Disk1 -quiet 1 -mailto -compress lzo -mode stop -mailnotification failure Unfortunately, if the backup tool is not solved (unreliable) it affects the whole infrastructure of my client. I hope it could be fixed, as it is pretty annoying not to have a robust backup system in which my company can trust.
Please feel free to ask any information which can help you to resolve this problem. We prefer using "stop" mode to garantee datas integrity during backup (important datas, with live production database running, etc.).
We use only enterprise channel to update (using subscription) to be sure to keep maximum stability. No problem on our proxmox instance, all work greats except this backup problem.
Our promox is fairly new (fresh install a few month ago as Proxmox 5-1), we upgraded recently to 5-2.5.
This can happen on any VM, those with Debian(+ pve guest agent) and those without guest agent (like a pfSense instance).Īll VM restart ok (i think this is what the patch has corrected), but the backup itself is not done. Yesterday we get a fail on 2 VM on the 6. We backup the same 6 VMs each days (some of them pretty big), and around 2 times per weeks we have this error, making the backup for the VM fail. INFO: vm is online again after 22 secondsĮRROR: Backup of VM 100 failed - start failed: : Unit 100.scope already exists. INFO: starting kvm to execute backup task That is why i open this issue separatly, to be specific on the problem of the fail of the backup. Now the VM reboot always ok with this patch, this is great, but it seems there is always the problem of failed backup. I read on other thread there was a problem and said it has been fixed since (please see: ). I would like to report there is always the problem of random failed scheduled backup, in "stop" mode, on VMs on Proxmox 5.2-5.