AAnonymous2d ago · 4 min readYour System Design Looks Fine on a Whiteboard. Until Traffic Hits.You can draw a system with an API gateway, a few services, Redis, and a database. Everything looks clean. Every component has a purpose. But what happens when traffic increases tenfold? Does Redis prostackmonkey.hashnode.dev
SSui Gn5d ago · 6 min readThe Holy Grail of Autonomous Cognitive Computing: Breaking Contextual Momentum in LLMs Large language models generate responses conditioned on the context they receive—including their own previous outputs. An early assumption can become the premise for later reasoning, allowing a mistakneurons-me.hashnode.dev
HHamed Yousefi6d ago · 10 min readCutting etcd/raft failover from 1.25 s to 343 ms, checked against 2,000 simulated leader crashesWhen the leader of a three-node etcd/raft cluster dies, the other two servers wait between 1 and 1.9 seconds before one of them holds an election. That's with a tick every 100 ms and the settings fromhmdsefi.hashnode.dev
YYash Suthar6d ago · 4 min readHow Redis rate limiters fail under load, measuredResearch paper: Overshoot, Retries and Clocks: An Empirical Study of Redis-Based Rate Limiting Under Concurrency · code and data · DOI 10.5281/zenodo.23467402 I build rate limiting for a living, and yashrsuthar.hashnode.dev
MMadhav BhasinOct 16 · 7 min readWhat I Learned Running Real-Time Pipelines at ScaleFor the past few years I've spent a significant portion of my time building and operating real-time event pipelines — systems that ingest millions of messages daily, process them through multiple serv47JDYblog.madhav.dev
MMukul KumarOct 17 · 4 min readHow Discord’s Message Storage Hit a Scaling WallDiscord’s migration from Cassandra to ScyllaDB is easy to summarise as a database replacement. That misses an important part of the engineering story. Discord reported reducing its Cassandra fleet fro2Sbuildopsy.hashnode.dev
RRajat RaiOct 15 · 17 min readWhen Retrying Makes Things WorseWhenever a request fails, one of the first things that comes to our mind is to retry it. The request failed? Try again. The API timed out? Send it again. Maybe it will work this time. And honestly, th13Mrajatrai.hashnode.dev
PPatrick KearnsOct 13 · 14 min readYour Distributed Lock Expired. Your Code Didn't.A distributed lock can expire while the code that acquired it is still running. Another instance takes ownership, finishes its work and saves the result. Then the original instance resumes and overwri911DBMfullstackcity.com
HHamed YousefiOct 13 · 9 min readA Go test that passed 40,000 times, then lost a write under loadSay a client writes a value and gets a success back. Later, the value is gone. It happens once in production, around the time a node restarts, and every test is green. A week later a CI job fails on ahmdsefi.hashnode.dev
PParminder singhOct 13 · 12 min readI Thought I Knew What "Scalable" Meant. Six Interview Questions Proved Me Wrong.Every system design resource I've opened throws around the same handful of words. Scalable. Highly available. Throughput. Stateless. I've used all of them in standups for years, so I assumed I understsystemdesignwithparam.hashnode.dev
