Today’s Heavy Networking dives into QUIC, a transport protocol originally developed by Google and now an IETF project.

QUIC promises advances over TCP and the opportunity for faster innovation, but it also poses a potential problem; developers may try to tune congestion control algorithms such as BBR in their favor, gobbling up available throughput while other connections limp along.

If competitors respond with their own tweaks, this could, as our guest writes “trigger a race to the bottom, and potentially break the Internet, unless network providers start deploying adequate Active Queue Management algorithms.”

That guest is Christian Huitema, an author and contributor to IETF standards with more than 40 years’ experience with software and Internet protocols. A former Microsoft engineer, he’s founder of the consulting firm Private Octopus, Inc.

We discuss how QUICK and BBR function, how they work around traditional gatekeepers and the pros and cons of those workarounds, and how QUIC differs from TCP.

We also examine the potential problems that QUIC could unleash, and explore congestion control and Adaptive Queue Management (AQM) algorithms as possible solutions.

Show Links:

Will Transport Innovation Collapse The Internet? – Christian Huitema

Christian Huitema’s Blog

Christian Huitema on LinkedIn

QUIC

BBR

Private Octopus, Inc.

QUIC, a multiplexed stream transport over UDP – The Chromium Projects

Network scheduler – Wikipedia

RFC 7567 – IETF Recommendations Regarding Active Queue Management – IETF

BBR, the new kid on the TCP block – APNIC Blog

TCP and BBR (2018) – Geoff Huston (PDF)