May 5 · 4 min read · Early in my time on the Kubernetes team, a customer proposed something that was both brilliant and beyond what we were ready for: a global footprint of clusters, one per region, with a synchronized set of jobs. They were running a low-latency applica...
Join discussionMay 3 · 20 min read · TLDR: Split brain happens when a network partition causes two nodes to simultaneously believe they are the leader — each accepting writes the other never sees. Prevent it with quorum consensus (at least ⌊N/2⌋+1 nodes must agree before leadership is g...
Join discussionMar 9 · 13 min read · TLDR TLDR: Consensus algorithms allow a cluster of computers to agree on a single value (e.g., "Who is the leader?"). Paxos is the academic standard — correct but notoriously hard to understand. Raft is the practical standard — designed for understa...
Join discussionDec 27, 2025 · 4 min read · 💡 I have tried an implementation of Raft, following the official Raft paper. Check the complete implementation here - https://github.com/iamvineettiwari/raft. This is a brief blog intended to provide a quick overview of Raft’s core components and ho...
Join discussionNov 16, 2025 · 18 min read · Introduction to Etcd Hey everyone! After our successful blog on k8s architecture, we're now going deeper into each component of the k8s architecture. In this blog, we'll focus on etcd - the brain of a k8s cluster and one of the most important topics....
Join discussion
Aug 31, 2025 · 2 min read · 1. 📈 Eventual Consistency (Weakest) Key Principle: All nodes will eventually converge to the same state, but not immediately. ✨ Characteristics: Finite Time Convergence: All nodes will eventually reach the same state after a finite period No Immed...
Join discussionJul 28, 2025 · 3 min read · Do you know you can run multiple Consul servers on a single node to quickly set up a Consul cluster for testing, eliminating the need for separate VMs? To run a Consul server on a single node, use the following command: ~ consul agent -server \ -no...
Join discussion
Jul 21, 2025 · 14 min read · It was a Tuesday, the kind of unremarkable day that precedes most production fires. The team, sharp and capable, had built a new distributed job scheduling service. To handle failover, they implemented what seemed like a clever, simple leader electio...
Join discussionMay 31, 2025 · 6 min read · Desde hace tiempo me han estado preguntando sobre este tema, así que vamos a crear un clúster de Hashicorp Vault, algo fundamental para una empresa moderna. Una bóveda que nos ayudará a gestionar mejor los secretos y también a reducir los costos de n...
Join discussion
