A framework for measuring Internet resilience

Cloudflare’s blog post outlines a data-driven framework for measuring Internet resilience that moves beyond uptime to capture a network’s ability to withstand, adapt to, and recover from disruptions. The framework relies on public data sources to quantify resilience across physical infrastructure, topology, interconnection, and routing behaviour, with the aim of making measurements reproducible for policymakers and operators alike.

It places emphasis on how local decisions and diverse interconnections shape national and regional resilience within the global Internet, recognising the decentralised nature of routing and the importance of multiple, independent paths. Route hygiene is a core theme, highlighting measures such as RPKI and BGP route validation, plus the role of IXPs, multi-homing, and diverse submarine cables in reducing single points of failure.

The post notes that measuring resilience is challenging because detailed network maps are sensitive, so the approach blends partial perspectives from tools like BGP data, traceroute, and IXP maps, with impact-weighted metrics to reflect real user exposure. It also references historic outages, such as the 8 July 2022 Rogers incident affecting over 12 million users, to illustrate why resilience matters.