Dev Meeting 001: Kubernetes is a Framework
My notes on Kubernetes as a framework: from Google's scaling challenges and Borg to Operators, CRDs, and modern container orchestration basics.
Key takeaways
- Kubernetes evolved from Google's internal Borg system after 15+ years of production use.
- It is a framework, not just a tool. CRDs and Operators extend it to any workload.
- Self-healing: auto-restarts failed containers and reschedules workloads across nodes.
- CNCF governs Kubernetes and the broader cloud-native ecosystem.
What is Kubernetes?
Kubernetes (K8s) is a portable, extensible, open-source platform for managing containerised workloads and services. It supports declarative configuration and automation. Google open-sourced the project in 2014, building on over 15 years of experience running production workloads at scale.
The name originates from Greek, meaning "helmsman" or "pilot." K8s is shorthand — counting the eight letters between "K" and "s."
Automates deployment, scaling, and management of containerised applications across clusters.
Define desired state in YAML — Kubernetes ensures the cluster matches it.
Restarts failed containers, replaces nodes, and reschedules workloads automatically.
Google's Problem with Scale
- Google ran millions of data jobs across datacenters — for search, indexing, and analytics
- As the business expanded, managing multiple tenants and diverse workloads became increasingly difficult
- They needed a standard way to schedule resources automatically at massive scale
This led to Borg, Google's internal cluster manager — the direct predecessor to Kubernetes.
How Kubernetes Was Born
- Google extracted its resource management approach from Borg into an open-source project
- Together with Intel and Red Hat, they founded the Cloud Native Computing Foundation (CNCF)
- CNCF became the governance body standardising container orchestration and the broader cloud-native ecosystem
- Kubernetes quickly became the industry standard for deploying and managing containers
Kubernetes is not just a tool — it's a framework. Through Operators and Custom Resource Definitions (CRDs), teams can extend it to manage any type of workload, from databases to ML pipelines.
Summary
Many thanks to Marcin Karkocha for presenting on Kubernetes at our Dev Meeting, the internal series I ran for my team at the time. The live demo showed theory in action — deploying and scaling containers in real time.
Dev Meeting 004: Introduction to Event Storming
My notes on Event Storming, a workshop-style method for exploring complex systems and business domains by mapping the flow of events and interactions.
Did You Know? 10 Developer Tips from Real Codebases
A collection of practical software development tips, workarounds, and clever solutions discovered in production codebases. Part 1 of an ongoing series.