Same here. I used proxmox for 8 years and have recently dumped it in favour of a couple of incus machines running OCI and LXC containers.
Much lighter, much faster, and to be honest, more straightforward when it comes to storage abstraction, which I think proxmox does in a very… convoluted way.
You can create a canonical name (CNAME) record to point to your old domain, you don’t need to recreate the LE if you don’t want to.
Assuming I understood your scenario, Canonical name record pointing one DNS name to another domain, then recreate your letsencrypt on the host.
“18% of car owners don’t know their brake fluid DOT rating.”
Makes sense, thank you.
Looks great, and I’m fully supportive of this.
However, I can’t understand the use case, and the part I can’t wrap my head around is why the server part is required… What differentiates pinepods from, say, antennapod from a functional standpoint?
I thought I was being clear that I have audited some of the scripts. They are built referencing other scripts instead of functions, and these rely on URLs. It’s difficult to follow.
Don’t ask chatgpt to audit code.
Have you ever looked at what was once ttek scripts? They’re a spaghetti of calls to other scripts. It’s not pretty. And not intuitive to audit.
An apipa address is a sign that networking is not working as intended. This should be resolved first before assigning a class C private addr manually.
deleted by creator
The ds211j is on synology DSM 6, which is ancient. I’ll look again, but I don’t think it supports btrfs.
I don’t use access control, I lock down with networking and filters.
But those are concerns for all containers, and not really an immich issue.
I agree, NFS is eazy peazy, livin greazy.
I have an old ds211j synology for backup. I just can’t bring myself to replace it, it still works. However, it doesn’t support zfs. I wish I could get another Linux running on this thing.
However, NFS does work on it and is so simple and easy to lock down, it works in a ton of corner cases like mine.
deleted by creator
My issue with mgmt.config is that it bills itself as an api-driven “modern” orchestrator, but as soon as you don’t have systemd on clients, it becomes insanely complicated to blast out simple changes.
Mgmt.config also claims to be “easy”, but you have to learn MCL’s weird syntax, which the issue I have with chef and its use of ruby.
Yes, ansible is relatively simple, but it runs on anything (including being supported on actual arm64) and I daresay that layering roles and modules makes ansible quite powerful.
It’s kind of like nagios… Nagios sucks. But it has such a massive library of monitoring tricks and tools that it will be around forever.
Running the k8s in their own VM will allow you to hedge against mistakes and keep some separation between infra and kube.
I personally don’t use proxmox anymore, but I deploy with ansible and roles, not k8s anymore.
Besides being easy, there is no advantage to tailscale for this case, and I would add that lots of us don’t want to depend on an external resource just to road warrior back home.
I think you’re still OK.
There are lots of great itx boards out there, so basically take your pick. And there will be ppl who say you don’t need ecc, but it certainly won’t hurt, it will just cost a bunch more.
I run a supermicro mini itx with ecc, for my zfs backup, but my non ecc machines have also been fine.
That thing about docker being so badly behaved in unprivileged containers seems to be a proxmox problem, not an LXC problem, as I’ve discovered running LXC in a non-proxmox environment.