Ancestors

Toot

Written by pezhore on 2024-08-29 at 15:35

I was productive yesterday - I spun up my first kubernetes cluster on my three node Proxmox cluster. The controller and 3x worker nodes are Rocky LXC containers.

The problem I'm attempting to solve is this monolithic beast of a docker host VM - it's running 12 containers and is struggling. I'm hoping to move to kubes to

[#]kubernetes #proxmox #lxc #learnk8s #homelab

=> View attached media

=> More informations about this toot | More toots from pezhore@infosec.exchange

Descendants

Written by pezhore on 2024-08-29 at 17:32

And with some more trickery and volume mounts, I have pdns working for an external-dns endpoint.

Things are coming along smoothly, I think i'm ready to deploy invidious on kubes.

[#]invidious #kubes #homelab #selfhosted

=> More informations about this toot | More toots from pezhore@infosec.exchange

Written by Michael on 2024-08-30 at 20:20

@pezhore That sounds like you got a lot of the basic infra in place. 👍

Also one great advantage of a workload scheduler and more than one physical node: You can just shut down any one of the nodes for maintenance and everything automagically stay up, with only a minimal interruption.

=> More informations about this toot | More toots from mmeier@social.mei-home.net

Written by pezhore on 2024-09-01 at 01:16

@mmeier I'm really looking forward to the resiliency aspect of this. Not sure how some containers would like having multiple replicas - but at least being able to move things around for scheduled outages would be nice.

=> More informations about this toot | More toots from pezhore@infosec.exchange