Reminder to self: see how widespread support for the TCP BBR congestion control algorithm is.
Posted by jpluimers on 2019/01/14
The TCP BBR congestion control algorithm was introduced in September 2016 and became available in Linux kernel 4.9 in July 2017 after being in the news for a good 5 months (see links below). It strives for better bandwidth use and lowering latency on big data pipes.
This post is a reminder myself to see how widespread that is on recent distributions for both end-user and server systems.
- [WayBack] google/bbr
- [Archive.is] kernel/git/davem/net-next.git – David Miller’s -next networking tree
- [WayBack] kernel 4.9: please enable TCP: BBR congestion control algorithm · Issue #1784 · raspberrypi/linux
- [WayBack] BBR: Congestion-Based Congestion Control | February 2017 | Communications of the ACM Measuring bottleneck bandwidth and round-trip propagation time.
- [WayBack] BBR, the new kid on the TCP block | APNIC Blog: A look at Google’s delay-controlled TCP flow control algorithm, called BBR.
- [WayBack] Google speeds up Cloud Platform with new networking algorithm | ZDNet The BBR (“Bottleneck Bandwidth and Round-trip propagation time”) algorithm does away with the assumption that packet loss equals network congestion.
- [WayBack] Google Cloud Platform Blog: TCP BBR congestion control comes to GCP – your Internet just got faster
- [WayBack] Google Cloud gets a new networking algorithm that boosts internet throughput | TechCrunch
- [WayBack] Increase your Linux server Internet speed with TCP BBR congestion control – nixCraft
- [WayBack] BbrTcp < PERTKB < TWiki
Via:
- [WayBack] With this simple config option you can push more data between long distance Linux server. I went from 200 Mbit/s to 800 Mbit/s. – nixCraft – Google+
- [WayBack] With this simple config option you can push more data between long distance Linux servers … (where Kim Nilsson mentions:Wow, just found this.
“The 4.10 kernel will also become the LTS Rolling Kernel in 16.04 LTS” coming with 16.04.3 in August
So, not such a long wait after all. :-)
–jeroen






Leave a comment