Tech

What’s Crossplane? Crossplane is an Open Source tool, which allows you to manage Cloud infrastructure as Kubernetes objects. In other words, you can create, modify, and delete AWS cloud assets using only Kubernetes manifests in the same way as with Terraform (or another IaC tool). It also allows you to manage cloud resources in different public clouds using a concept of providers very similarly to Terraform. Therefore, Crossplane allows engineers to manage the whole application lifecycle from a single entry point e.g. a Helm chart. ...

I’m writing a post on NeoVim and NeoVim GUI configuration mostly for myself. Even though I have all my dotfiles on the Github, some configuration requires gluing. Also, some things have to be executed in a particular order and some configs must exist in particular places. I spent some time looking for fixes for occasional issues and polishing some rough edges. So, here I just want to put it all together, so I won’t need to go down the googling spiral if I need to set up a new development environment for myself. ...

Pulumi is just another Infrastructure as Code tool, yet it’s different from what I’ve seen before. It offers some interesting concepts like usage of the generic purpose programming languages, state tracking, etc. I had some time recently so I decided to play around with it. Here I want to share some thoughts about this tool. In my opinion this is the first IaC tool, which was created for the “Dev” side of “DevOps”. With the raise of DevOps movement, both software and system engineers learned a lot from each other. Infrastructure as Code itself is a SWE approach. However, infrastructure usually falls into the Ops people responsibilities. Therefore, IaC tools as well as configuration management tools were built to be convenient for sysadmins. We had several different DSLs and tons of YAML. Pulumi is different. It doesn’t re-invent ways to declare your infrastructure with code. ...

Image source I’ve been using Kubernetes for almost a year and want to share some takeaways and things I’ve learned about it so far. Some context. We had a team of two engineers working not only on Kubernetes but on all infrastructure-related topics. Our main application was a Python monolith, plus a few satellite-services. Our infrastructure was dynamic due to autoscaling, and a lot of things were optimized for costs (check out my article about running Kubernetes on AWS Spot instances). ...