Little bit of everything!

Avid Swiftie (come join us at !taylorswift@poptalk.scrubbles.tech )

Gaming (Mass Effect, Witcher, and too much Satisfactory)

Sci-fi

I live for 90s TV sitcoms

  • 4 Posts
  • 40 Comments
Joined 2 years ago
cake
Cake day: June 2nd, 2023

help-circle
  • Not at all. Proxmox does a great job at hosting VMs and giving a control plane for them - but it does not do containers well. LXCs are a thing, and it hosts those - but never try to do docker in an LXC. (I tried so many different ways and guides and there were just too many caveats, and you end up always essentially giving root access to your containers, so it’s not great anyway). I’d like to see proxmox offer some sort of docker-first approach will it will manage volumes at the proxmox level, but they don’t seem concerned with that, and honestly if you’re doing that then you’re nearing kubernetes anyway.

    Which is what I ended up doing - k3s on proxmox VMs. Proxmox handles the instances themselves, spins up a VM on each host to run k3s, and then I run k3s from within there. Same paradigm as the major cloud providers. GKE, AKS, and EKS all run k8s within a VM on their existing compute stack, so this fits right in.



  • I think at this point I agree with the other commenter. If you’re strapped for storage it’s time to leave Synology behind, but it sounds more like it’s time to separate your app server from your storage server.

    I use proxmox, and it was my primary when I got started with the same thing. I recommend build out storage in proxmox directly, that will be for VM images and container volumes. Then utilize regular backups to your Synology box. That way you have hot storage for drives and running things, cold storage for backups.

    Then, inside your vms and containers you can mount things like media and other items from your Synology.

    For you, I would recommend proxmox, then on top of that a big VM for running docker containers. In that VM you have all of your mounts from Synology into that VM, like Jellyfin stuff, and you pass those mounts into docker.

    If you ever find yourself needing to stretch beyond the one box, then you can think about kubernetes or something, but I think that would be a good jump for now.







  • There’s many ways to do this. Saving the disk state is one, I believe that’s what the other person suggested - essentially stores the disk as an image which then you use for future vms as your jumping off point. This is also essentially how workstations are deployed at companies. (Essentially being the key word). Cloud providers have different names for this too, in AWS this is called their AMI.

    Another option is Ansible, which essentially handles deploying a VM by running your scripts for you. I haven’t played too much with this, and I doubt it works with VirtualBox, but it’s something you may want to look into, it would definitely uplevel your skills.

    Thirdly is dependent on what you actually use your VM for, you haven’t given your use cases but this is one of the reasons containerization became such a thing - because when running an app we mostly don’t care about the underlying system. It may be worth it to learn about docker.










  • Pixelfed won’t launch groups anytime soon unfortunately, I’ve followed closely and it seems it’ll be out about the same time as Tesla’s full self driving.

    I think the best way would be choose yourself piefed or Lemmy, and then you offer mastodon or pixelfed as places for your users to join and subscribe. So it’s a little more natural to them.

    While your goal is noble, reality is that I would expect a very small percentage of people to migrate. Less than one maybe even. I think that even a few hundred more people joining the fediverse here is a great idea, but I don’t want you to be discouraged when it doesn’t turn out the way you hope. People love their platforms. (I also just read your last paragraph after writing this, my bad, glad you’re in the right headspace :)