Free Uptime Monitoring Insights

Insights, comparisons, and best practices on website monitoring, uptime checking, and site reliability.

Press Esc to clear

Showing 12 of 142 articles

Founder Log

May 22, 2026•11 min read•

GitHub Was Breached Through The Front Door

The GitHub internal repository breach was not a freak accident. It was the obvious result of treating developer machines, editor extensions, and marketplace badges as trusted infrastructure.

Incident Management

May 22, 2026•7 min read•

Morten Pradsgaard

How to Watch DNS, TCP, TLS and TTFB Change in Real Time While Debugging a Slow Endpoint

When an API gets mysteriously slow, the answer is hiding in the phase breakdown. Here's how to actually see DNS, Connect, TLS, and TTFB shift live during an incident.

Server & Infrastructure Monitoring

May 22, 2026•9 min read•

Morten Pradsgaard

WebSockets vs Polling for Monitoring Dashboards: Why We Stream 30-Byte Probes Instead of Refreshing

An engineering breakdown of why polling is the wrong default for a probe stream — and how we built a WebSocket transport that paints in sub-second time with a 30-byte wire format.

Website Monitoring

May 22, 2026•6 min read•

Morten Pradsgaard

Most 'Real-Time' Uptime Monitors Refresh Every 30 Seconds. Here's What Actually Live Looks Like.

Pingdom, UptimeRobot, Better Stack, Checkly, Hyperping, StatusCake — every 'real-time' dashboard is just a polled API on a 30–60s refresh timer. There's another way.

Founder Log

May 1, 2026•6 min read•

Morten Pradsgaard

GitHub Doesn't Get To Decide What 'Down' Means

GitHub has had three months of corrupted merges, broken search, and auth outages — and the status page is still mostly green. That's not because the platform is healthy. It's because GitHub gets to define the word 'down,' and they've defined it generously.

Domain Intelligence

April 12, 2026•4 min read•

Morten Pradsgaard

Catching DNS Hijacks Without Crying Wolf

Most uptime monitors check whether a site responds. Fewer check whether the site responding is actually yours. Here's the simple pattern we use to detect DNS hijacks without alerting on every legitimate record change.

Founder Log

April 12, 2026•6 min read•

Morten Pradsgaard

Engineering Internals: A 20-Line Semaphore and a Scheduler That Barely Touches the Database

Two small but load-bearing pieces of code from our check runner: a lock-free semaphore that fits on a napkin, and an in-memory schedule that cut our database reads by over 99%.

Founder Log

April 12, 2026•5 min read•

Morten Pradsgaard

The Night We Almost Paged 10,000 People Because Our Own VPS Blinked

When a thousand unrelated sites all fail in the same second, the problem is almost never a thousand unrelated sites. Here's how we built a circuit breaker for our own alerts — and why every monitoring system needs one.

Domain Intelligence

April 9, 2026•5 min read•

Morten Pradsgaard