VM redundancy
VM redundancy
I had proxmox with unattended updates in VMs, but VMs sometimes hangstop booting, encounter DNS issues, etc. That’s not something that I can rely on!
I want to run a vm and make sure the service in it is always available. The users should always have access to the service through Cloudflare tunnels. The bare metal installation works for a long time, but not VM installation. I can tolerate a 30 minutes fix once year, but it won’t work telling users sorry there is a problem this week with my server!!!
How to set up this VM?
So far, the solutions that I found are, set up automatic ZFS snapshots every 15 minutes, and set up another proxmox server with proxmox backup server, ready to boot from backup.
Not sure what you expect. Hangstop booting and DNS issues is so vague, could you provide more details? What happens exactly and how do you fix it usually? Any logs you can provide? What OS is running on the VMs?
1Yeah. Unless I purposefully or accidentally mess something up, my VMs run great without interaction. But either way, I usually have to manually change something to create an unstable vm.
-
Check your host syslog (it’s available in the GUI) and it should give you an idea of why the VMs are hanging.
-
Have you considered a high availability cluster? It’s literally the tool that exists for what you want.
PBS is cool for like a disaster recovery scenario, but it shouldn’t be the tool you go to for frequently unstable VMs. That’d be a fabric bandaid at best, but leaning heavily to a liquid bandaid in practice.
1-
I suppose you could deploy an HA cluster using either Ceph or StarWind VSAN HA storage technology. These solutions replicate data across cluster nodes, ensuring your VMs are highly available. They both fit Proxmox.
1