Field notes
When a CDN, a cloud region or a DNS provider fails, many of the sites that depend on it fail at the same time, and their owners have to work out quickly whether the fault is theirs. This category looks at real incidents of that kind: Cloudflare in November 2025, AWS us-east-1 in October 2025, Fastly in June 2021 and Facebook in October 2021, among others.
Most posts work from the provider's own published post-mortem, and each one ends with the monitoring lesson: what an external check would have caught, what it could not see, and which layers, such as DNS, are worth checking separately from the page itself. If you want one post, read the first in the list below. It covers four incidents and ends with a short checklist for the hours after an outage. The other two go deeper on CDN failures.
Showing 3 articles. Page 1 of 1.
Fastly's June 2021 outage and AWS US-EAST-1 DNS failures in Q4 2025 show how infrastructure dependencies can disrupt websites. This analysis explores the causes, operational impacts and role of monitoring in incident response.
Four documented outages, from Cloudflare, AWS, Fastly and Facebook, and the check that would have shown each one.
What broke in Cloudflare's 18 November 2025 and 20 February 2026 outages and AWS's us-east-1 failure of 19 to 20 October 2025, how long each lasted, and which monitoring layers catch them.
Your first check is free when you add your site. Then try a paid plan for 7 days (card required) or subscribe, from $15 USD a month.
Start 7-day trial