Maybe a potential option is Opnsense on some x86 hardware? Would be really easy to find a mini PC or something.

Interesting, waiting on network mounts could be useful!

I deploy everything through Komodo so it’s handling the initial start of the stack, updates, logs, etc…

That’s wild! What advantage do you get from it, or is it just because you can for fun?

Also I’ve never seen a service created for each docker stack like that before…

Doing a volume like the default Immich docker-compose uses should work fine, even through restarts. I’m not sure why your setup is blowing up the volume.

Normally volumes are only removed if there is no running container associated with it, and you manually run docker volume prune

Do a one-time copy of your photos using SFTP or FolderSync or whatever works for you.

Well one advantage is you can just use this tool instead of needing to use each website or app individually, it saves a lot of time that way.

Also might as well have one less company collecting data on you, vs using a commercial closed source social media management tool.

Batch mode is great! I didn’t realize it was added until just now hah

https://github.com/kd2org/karadav

Nextcloud client/app compatible WebDAV server with a lightweight file browser webUI, and multi-user support.

Should be the closest thing to Google Drive without actually running Nextcloud.

The only issue is it looks like the Nextcloud iOS clients don’t work.

It’s pretty easy, you can browse files in an LXC backup and restore specific parts. For VMs you can just restore the whole VM and copy out what you need.

I back up all the directories and docker-compose files using Restic (via Backrest) stored on Backblaze B2, and also the whole Docker LXC via Proxmox’s backup function to a local HDD.

There’s a chance some databases could be backed up in an unusable state, but I keep like 30-50 snapshots going back months, so I figure if the latest one has a bad DB backup, I could go back another day and try that one.

I also don’t really have irreplaceable data stored in DBs, stuff like Immich has data in a DB that would be annoying to lose, but the photos themselves are just on the filesystem.

For testing Restic I pull a backup and just go through and check some of the important files.

Proxmox backup is really easy to test, as it just restores the whole LXC with a new ID and IP that I can test.

NP! That’s how I do it on proxmox, I’ll start the VM every so often and update it. Only takes a few seconds to clone so it’s nice and quick to do.

Simple method is just keep a ready to go VM and clone it.

What’s the difference vs using a governer with scaling like ondemand?

You can either:

A) Use a different port, just set up the new service to run on a port that’s not used by the other service.

B) If it’s a TCP service use a reverse proxy and a subdomain.

It’s just a YAML thing, if you do FILEBROWSER_CONFIG:"/config/config.yaml" instead it might work with quotes.

It’s interesting because you’re not the first person to complain about getting ISOs in Proxmox, but on my instance if I click on my local storage it has an upload ISO button, and a download ISO from URL button right there, so it’s really simple.

It can also mount network storage with existing ISOs and just pull from that.

I don’t use ISOs very often though, either a Debian 12 container template, or a custom Debian 12 cloud-init VM I made and backed up, so I can just hit restore and it gives me a fresh VM with new networking config and everything through cloud-init automatically.

Is it all automated with versioning intervals and stuff? Or is restic required as a third party step and maintaining a duplicate of data on the server for it to grab?

Overall it sounds like a decent VM manager but is meant for enterprise stuff where they’ll be building their own backup systems.

That’s what proxmox has too, but snapshots aren’t backups and aren’t being sent to a remote backup server… You’re also not supposed to keep snapshots around for very long, whereas I have backups going back several months.

Or are you sending snapshots to a remote server? I think ZFS can do that, so maybe that’s an option I can look at.

This looks interesting, how do you handle automated backups of all the VMs/Containers? Their docs kind of seem to say “stop everything and figure it out”, but with Proxmox I’m used to it handling everything automatically to my PBS server every night.

It’s definitely not an easy migration in my experience, because they run rootless and they cannot auto-start without making a system service for every stack, there is a lot that needs to change in a compose stack, especially with file permissions for shared mounts.