Guides to downtime, errors and monitoring

Five subjects, each with a hub page that frames the problem and a set of articles that go into it properly. Written to be read once and remembered, not to be skimmed for a keyword.

Major outage history

Detailed accounts of the internet's largest outages - What broke, why it cascaded so far, and what the industry changed afterwards.

All 7 in major outage history

Where to start, depending on what just happened

A site will not load and you do not know why

Begin with is it down for everyone or just me, which settles the first question in about a minute, then run the site through our checker to see which stage of the request actually failed.

You have an error code on screen

Go straight to the error code guides. A 502 and a 504 look identical to a reader and mean different things about who can fix it, and the code narrows the cause more than anything else available to you.

Everything is broken, not one site

That is usually your own connection or your provider rather than the whole internet. The troubleshooting hub works through the causes in order of how often each one turns out to be the answer, starting with DNS.

You are trying to measure or promise uptime

Read uptime monitoring for how measurement actually works, and use the uptime calculator to see what a percentage in a contract really permits. Our own methodology sets out what our figures include and, more usefully, what they cannot see.

You want to know what a big outage looked like

The outage history hub reconstructs the largest incidents on record: What broke, what the public saw, and how long recovery actually took once the fix was known.