Nelm is a Helm 3 alternative. It is a Kubernetes deployment tool that manages Helm Charts and deploys them to Kubernetes. It is also the deployment engine of werf. Nelm can do (almost) everything that Helm does, but better, and even quite some on top of it.

At RudderStack, we decided to use PostgreSQL as our main streaming engine and queuing system instead of specialized tools like Apache Kafka. We picked PostgreSQL because it's flexible, reliable for transactions, and easier to debug. If you are curious to learn more about that decision, read the previous post about the rationale behind why we chose Postgres over Apache Kafka and the initial architectural patterns we employed. Over the past six years, this system has proven reliable and has scaled to handle 100,000 events per second—but only after we successfully navigated challenges like table bloat, query performance degradation, index bottlenecks, and retry storms.

This post is a chronicle of the critical, hard-won lessons learned while maturing PostgreSQL into a highly performant and resilient queuing system.

Terminal-based group chat app with real-time WebSocket messaging, file sharing, themes, and admin tools — built with Go and Bubble Tea.

Self-hosted task management that combines the simplicity of personal with the power of professional project organization. Built for individuals and teams who value privacy, control, and efficiency.

Have you heard of Kubernetes (also known as k8s)? Until a few months back, I knew it existed and that it was like infrastructure’s holy grail. It has to cover the basics, like auto-scaling and load balancing or automated rollbacks… And then there are millions of tools to build on top of it.

As we recently migrated our deployment from AWS Elastic Container Service (ECS) to AWS Elastic Kubernetes Service (EKS; managed Kubernetes cluster), I wanted to share some tips. It also feels nice to do that on the 10th anniversary of “Kubernetes: The Future of Cloud Hosting” MeteorHack’s blog post.

Please keep in mind that a Kubernetes cluster is an extremely complex beast, and I’m pretty far from being able to explain all the “whys” you may have. Our amazing DevOps Engineer managed to make it work, and I’m really happy with the current setup. Both because the app performs better at a lower cost and because I learned a lot along the way.

Kubermatic Kubernetes Platform is in an open source project to centrally manage the global automation of thousands of Kubernetes clusters across multicloud, on-prem and edge with unparalleled density and resilience.