Introduction It was 12 PM on a weekday - our peak traffic hour. I was one of the three engineers at a small fintech startup. Suddenly, our entire platform went dark. No metrics, no logs, no alerts - just angry customers and a silent monitoring dashboard. We were in limbo and no idea what failed. After 45 minutes of blind debugging and a full system restart, we were back online. That day taught me that fault tolerance isn’t luxury engineering - it’s survival....
The Quiet Beginning and the First Ripples In 2010, a Mozilla researcher named Graydon Hoare began working on a personal project born from frustration—a stuck elevator had triggered a software failure that could have been prevented with better memory management. This moment of annoyance would eventually give birth to Rust, a language that would fundamentally challenge how we think about systems programming. For its first few years, Rust remained largely within Mozilla’s walls, quietly evolving through rapid iterations....
Database Design: A Product-First Approach to Building Scalable Systems Over the years wrestling with database architectures I’ve learned that the most elegant technical solution isn’t always the right one. Database design is often viewed through the lens of technical optimization—normalized tables, efficient indexes, and query performance. While these technical aspects are crucial, the database disasters I’ve witnessed taught me that the most successful architectures emerge from a deep understanding of the product they serve....
The Backstory I have been a vim user since a decade back and switched to using neovim more than five years ago. Anyone familiar with the vim/neovim ecosystem would know how much it has evolved in the last few years. I had started with using Vundle for managing my plugins, moving to Plug at a later point. All this while my configuration was a simple vim script (init.vim) where my configurations were neatly organized in less than ~200 lines....
In this blog we will learn about how to get started with Golang and create a auth with Google Login. We will be using the Gin and gorm ORM. During my initial uses of Gin I felt for new users the documentation is not friendly enough, hence I decided to write this article. If you are coming from a Python/Ruby world Gin is NOT like Django or Rails. Rather, its more like Flask and Sinatra....
In my previous post I wrote about process of pre deployment. A common question was about the deployment process itself. And the idea of this blog was born. The following is a non exhaustive common list of ways to deploy to cloud. In a single VM instance. For eg. Droplet(Digital Ocean), EC2(AWS), Compute Engine(GCP), etc. In one or more VM instances with autoscaling. Not all cloud providers currently support this. AWS, GCP and Azure currently does it....
Your app is built and you need to deploy it. What Next? Read On... Before we deploy applications we need to follow a set of process to ensure our application stack is prepared for any expected or unexpected events. Over the years I have experienced these scenarios and I have created a set of my personal checklists which allow me to quickly check that my application has all the requirements for recovering from potential pitfalls....