‹ Infrastructure, operations and security

A focused newsletter on cloud outages and their post-mortems, for on-call engineers who learn from other people's nights

When something happens

When something falls over: what broke, how long it lasted, and the provider's own post-mortem once it lands, plus what the status page's history shows over time. Nothing between incidents; for an on-call engineer who wants to learn from other people's nights, not a running uptime tracker.

Not minor blips with no post-mortem and no lasting lesson, not outage speculation before any official statement, not marketing framed as a reliability update, not uptime percentage league tables.

Where the news comes from: Cloudflare status history, AWS status, Google Cloud status, GitHub status history, DigitalOcean status history.

Nobody gets this one yet. It starts when you do.