Thoughts from IETF 109
Geoff Huston shares his thoughts from IETF 109, including on AS Prepending, Community Networks, DNS centrality, IPv6, and TCP.
Tag: TCP
Control congestion, lower latency with new L4S AQM scheduler
Guest Post: Virtual-Dual Queue Core-Stateless AQM is a new L4S compatible AQM that does not assume compatible TCP versions in any queue.
Modelling BBR’s interactions with loss-based congestion control
Guest Post: While BBR is supposed to be a rate-based algorithm, when competing with other flows, BBR is window-limited.
Bufferbloat may be solved, but it’s not over yet
Dave Täht recounted the story of bufferbloat and the deployment challenges ahead.
Sizing the buffer
If buffers are generally good and improve data throughput, then more (or larger) buffers are better, right?
BBR evaluation at a large CDN
Guest Post: Verizon Media shares its experience using BBR to deliver traffic from its caches.
Does TCP keep pace with QUIC?
Guest Post: Researchers have set out to compare the performance of an optimized TCP stack against QUIC.
A quick look at QUIC
Get the backstory on the history and design of QUIC, and what implications it might have on the overall architecture of the Internet Protocol.
It’s time to replace TCP with online-learning congestion control
Guest Post: Performance-oriented Congestion Control (PCC) is a new online learning approach to congestion control that seeks to overcome the limitations of traditional TCP.
BBR patches for higher WiFi throughput and lower queuing delays
Feedback is sought on recent BBR patches that seek to increase throughput for WiFi and lower queuing delays.