I'm a DevOps engineer (or SRE if you wish) with 5++ years of hands-on experience with a decent amount of tools that are most probably used or going to be used in your company. One of the most important tools that I love working with and want to continue working with, is Kubernetes. At least, while I don't see any better alternative to it. I think that containers themselves are one of coolest inventions in development, and I'm trying to use them as long as it's possible. Also, I believe that every routine must be automated, because routing is a boring job that lets people lose focus and make mistakes.
I think that there are several things that a good SRE or DevOps engineer must be able to do:
- To build reliable and stable infrastructure
- Keep this infrastructure up-to-date
- Keep all the source and instructions of this infrastructure clean and simple
- Avoid a human factor as long as possible
- And when it's not possible to avoid it, not to be afraid to take responsibility
Also, I think it's important that before implementing anything, an engineer has understood all the requirements and checked tools that can fulfil them. I often see, how people try to use a tool for its name but not for its functionality, and hence they have to do a lot of additional work and deal with compromises. And if nothing really can fulfil those requirements, you need not be afraid of writing something new *and open-source it*.
- I love to work with `Kubernetes`, but not with `yaml`.
- I'm a huge fan of [Helmfile](https://github.com/helmfile/helmfile).
- I have written several small cli tools in Rust, that you might find in my [GitHub profile pins](https://github.com/allanger) (they are not perfect, but I'm working on it).
- I'm trying to automate everything until I'm losing control over something that is automated.
- I love Perl, although I don't even remember how to write code with it, but I would be somehow thrilled to have any ability to work with it in production
- I also think that everything is better in Rust, or at least in Go *(if Bash is not enough)*
I have a blog (written-as-code) that is deployed to K8s (https://badhouseplants.net/), with the source code stored in a self-hosted Gitea, that is also deployed to K8s alongside the CI/CD system where this blog is being built and published. This CV is also being built as a part of the CI process, and then uploaded to `minio` storage that is also ~~surprisingly~~ running in this cluster. So you can download the latest version of CV here: <https://s3.badhouseplants.net/public-download/n.rodionov.pdf>
> But I can't guarantee 100% availability because it's a one-node k8s, and sometimes I need to do a maintenance work
