ohh also, the whole “hiding your mrf policies is sus” thing is so dumb, showing it is something large resilient instances can do, but I’m not looking to put a target on my back, having that public is a privelege
@getimiskon @cafkafk oh, you mean it's not just a typo of MtF policy? 😝
Thanks @getimiskon
@cafkafk ohhhh. My instinct is there's no way I'd leave that public.
@cafkafk ehhhh yes and no. It's the more general solution for sure but it looked like more boilerplate and less nix integration. Unless...you get bored on break from eza...👀
@cafkafk apropos of nothing did I show you the cluster IPv6 plans?
@cafkafk https://arichtman.github.io/nix/#topsoil-kubernetes
Should be cool writing a Lil Rust DHCPv6 client that writes a kubelet drop-in config file
@cafkafk Thanks! ideally it makes adding nodes a dynamic breeze! I'm also thinking to set up netboot.xyz so no need to ever flash ISOs to USB.
The final piece I want is the custom cluster autoscaler to power on and off machines. As soon as the cluster is functional again I should be able to continue writing that. Feels really dirty to shell out from Rust but I took a swing at the SSH library and it was kinda too low level for what I need it for
@arichtman i kinda also wanna write a module that does scaling automagically with microvms or maybe even hybridcloud but I just don’t have that kind traffic when I’m not at dayjob lol
@cafkafk I don't either but the idea is that the worker nodes can power down scale-to-zero when I'm like sleeping or whatever. If you find anything good like a template project for gRPC with protobuf stubs please send it my way. I think I can hack together something.