What It Actually Takes to Network a Live System That's Never Allowed to Go Offline

This story was originally published on HackerNoon at: https://hackernoon.com/what-it-actually-takes-to-network-a-live-system-thats-never-allowed-to-go-offline.
Lessons from building transit networks that can never go offline: addressing as a 20-year decision, blast-radius switch design, and testing by breaking things.
Check more stories related to cybersecurity at: https://hackernoon.com/c/cybersecurity. You can also check exclusive content about #networking, #system-design, #site-reliability-engineering, #infrastructure, #resilience, #building-networks, #public-transit-system, #intercom, and more.

This story was written by: @savni. Learn more about this writer by checking @savni's about page, and for more stories, please visit hackernoon.com.

I've spent over a decade building networks for a public transit system that's almost never allowed to go offline — surveillance, public address, intercom, and emergency comms running while trains keep moving underneath. That single "no maintenance window" constraint reshapes every decision: addressing schemes become twenty-year commitments, switch configuration becomes a blast-radius problem, and the acceptance test becomes the real deliverable. Here's what mission-critical network engineering actually takes when "down" means a passenger presses a call button and nobody answers.