What Is The Internet? Partial Connectivity in the Internet Core
Guillermo Baltra, Tarang Saluja, Yuri Pradkin, John Heidemann

TL;DR
This paper offers a formal definition of the Internet's connectivity, introduces algorithms to measure partial connectivity phenomena like peninsulas and islands, and provides empirical data on their prevalence and causes.
Contribution
It presents a testable, conceptual definition of Internet connectivity and algorithms for measuring partial connectivity, validated through large-scale measurements.
Findings
Peninsulas of partial connectivity are as common as outages.
Most peninsula-time is due to a few long-lived events.
Routing transients account for 45% of peninsula events.
Abstract
"A collection of interconnected networks" defines what the Internet is, but not what it is not. Events threaten Internet fragmentation: politics suggest countries or ISPs may secede or be de-peered, disputes between ISPs result in persistent unreachability between their customers, and architectural changes risk breaking the "one" Internet. Understanding such threats benefits from a testable definition of what the Internet is and is not, enabling discussion and quantification of partial connectivity. We provide a conceptual definition giving an idealized asymptote of connectivity. It implies peninsulas of persistent, partial connectivity, and islands when one or more computers are partitioned from the main Internet. We provide algorithms to measure, operationally, the number, size, and duration of peninsulas and islands. We apply these algorithms in rigorous measurement from two…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsNetwork Traffic and Congestion Control · Advanced Optical Network Technologies · Software-Defined Networks and 5G
