KubeDee is a Certified Kubernetes environment that uses LXD and is aimed mostly for test environments.
I’m testing this at the moment with OpenEBS as storage, and it works well so far.
Nice, but i don’t like to run some sh code on my machine without know what it doing.
The code is very long and so difficult to read.
The main script is 350 lines, https://github.com/schu/kubedee/blob/master/kubedee
The auxiliary script is 1500 lines, https://github.com/schu/kubedee/blob/master/lib.bash
It does not look that complicated. It looks well-written to me, and the author is a member to several other free software projects.
In any case, you would run this on a VPS or in a VM.
There is that point, it is a bit of a monster script to follow, but then again what about compiling some C code you can’t follow, whats the difference, it could still be malicious. Do you only compile and run software you have read inside out and can follow exactly what it does, ie you would be able to program at that level.?
If you want to feel safe in your own tin-foil hat security bubble, then run CentOS, selinux, an ancient kernel, on a VM and don’t use containers at all.
As stated by Simos, just run and test this in a VM in your own server, or a cloud VPS. Thats what I have done.
You are right but for infrastructure i prefer something like configuration tool: terraform, saltstack, ansible, puppet.
Are universal accepted, easy to read and every function is well documented.
This is not a critic to the project (very nice) but i prefer some playbook or formula
The original author of Kubedee has archived the project.
This fork has some activity and retains most of the
original shell implementation, so I’ll probably switch to it for now.
I have created my own fork, but my intention
is to rewrite Kubedee in Raku to help me learn the Raku language. So if you want something
immediately useful, check out zer0def’s fork.
I still think that Kubedee is a great project, and really demonstrates LXD’s capabilities as
a platform for developing distributed system software locally.