How do you do automated deployments, though? I don't like using K8s for small stuff, but I am also extremely allergic to having to log on to a server to do anything. Dokku hits the sweet spot for me, but at work I would probably use Nomad instead.
Set your pod to pull image always and have entrypoint shell script that clones the repo, kill the pod so on restart you could get your code deployed.
You could run init container with Kaniko that pushes image to repo and then main container that pulls that back but for that you need to do kubectl rollout restart deploy <name>
If you are looking for pure CI/CD gitlab has awesome support or you could do Tekton or Argo. They can run on the same cluster.
What's wrong with logging in to a server? I love logging in to a server and tinkering with it. Sure, for those who operate fleets of hundreds it's not scalable, but for a few servers that's a pleasure.
The problem is when you need to duplicate that server or restore it due to some error, you have no idea what all the changes you made are.
Besides, it's additional hassle and a chance for things to go wrong, the way I have it set up now is that production gets a new deployment whenever something gets pushed to master and I don't have to do anything else.
But this is a solved problem since... well, at least since the beginning of internet. I managed 1000s of Linux & BSD systems over the past 25 years and I have scripts that do all that since the mid 90s that automate everything. I never install anything manual; if I have to do something new, I first write + test a script to do that remotely. Also, all this 'containerization' is not new; I have been using debootstrap/chroot since around that time as well. I run cleanly separated multiple versions of legacy (it is a bit scary how much time it takes to move something written early 2000s to a modern Linux version) + modern applications without any (hosting/reproducibility) issues since forever (in internet years anyway).
True; I learned many years ago that that is not a good plan. Although, I too, love it. But I use my X220 and Openpandora at home to satisfy that need. Those setups I could not reproduce if you paid me.
> The problem is when you need to duplicate that server or restore it due to some error, you have no idea what all the changes you made are.
A text file with some setup notes is enough for simple needs, or something like Ansible if its more complex. A lot of web apps aren't much more than some files, a database, and maybe a config file or three (all of which should be versioned and backed up).
Make backup of /etc and package list. Usually that's enough to quickly replicate a configuration. It's not like servers are crashing every day. I'm managing few servers for a last 5 years and I don't remember a single crash, they just work.
I'm logging into a server because i need to, not because its 'pleasure'.
I don't hate it but if you need to login to a server regularly because you need to do an apt upgrade, you should have enabled automatic security updates and not login every few days.
If your server runs full because of some logfiles or stuff, you should fix the underlying issue and not needing to login to a server.
You should trust your machines, independently if it is only one machine, 2, 3 or 100. You wanna be able to go on holiday and know your systems are stable, secure and doing their job.
And logging in also implies a snow flake. Doesn't matter as long as that machine runs and as long as you have not that many changes but k8s actually makes it very simple to finally have an abstraction layer for infrastructure.
Flux or Argo can help with this. The operator lives on your cluster, and ensures your cluster state matches a Git repo with all your configuration in it.
This is what I like to do. In my case, even the CI/CD host is just a systemd service I wrote.
The service just runs a script that uses netcat to listen on a special port that I also configured GitHub to send webhooks to, and processes the hook/deploys if appropriate.
Then when it's done, systemd restarts the script (it is set to always restart) and we're locked and loaded again. It's about 15 lines of shell script in total.
