Hosts Mauven MacLeod and Graham Falkner deliver a fiery rant about the recent AWS US East 1 DNS outage and what it reveals about our dependence on cloud services. In this episode, they unpack the outage's real-world impact — from Snapchat and Venmo outages to Philips Hue bulbs and automated litter boxes going dark — and share colourful personal anecdotes, including a navigation fail on a Loch Lomond walk and a high‑tech mattress that turns into an expensive paperweight when the cloud hiccups. The pair dig into the technical and cultural roots of the problem: DNS as an ageing single point of failure, the dangers of concentrating critical infrastructure in one region, cost‑cutting that sacrifices resilience, and the worrying effects of automation and staff churn. They discuss how small businesses, banks, gaming platforms, and everyday consumers all found themselves unable to process payments, take bookings, or even turn on a light due to a single regional fault. Mauven and Graham also examine the human side of outages — exhausted sysadmins, online threads that read like group therapy, and the blurred line between human operators and automated systems shipping production code. They mock the absurdity of smart devices that need the internet to perform basic functions, and contrast that with the resilience of simple, offline tech (their beloved vinyl collections make a cameo). Finally, the episode offers a clear call to action: rethink resilience. Topics covered include multi‑cloud and hybrid strategies, decentralisation, offline fallback modes or “stupid mode” for essential devices, and the need to prioritise technical debt and redundancy over short‑term savings. Expect sharp humour, practical frustrations, and a promise of tangible fixes and advice in the next episode — plus plenty of memes and sympathy for the folks keeping the lights on.

Ask this episode

“What did the council decide on item 42?” — answered with the speaker and the moment.

Transcript

Related