Both organizations and everyday users are always glad to hear the internet is speeding up. Thanks to Google’s engineers – or rather their new congestion control algorithm, the Google Cloud Platform can achieve higher bandwidths and lower latencies for internet traffic. We’re talking about the same algorithm that powers TCP (Transmission Control Protocol) traffic from google.com. According to Google’s blog post, the implementation of this algorithm sped up even YouTube traffic by 4 percent, and up to 14 percent in some countries. Given that the YouTube traffic is highly-optimized, these results are quite impressive.
How does this algorithm work?
“BBR (“Bottleneck Bandwidth and Round-trip propagation time”) is a new congestion control algorithm developed at Google. Congestion control algorithms — running inside every computer, phone or tablet connected to a network — that decide how fast to send data,” explains Google in the blog post.
Since the late 1980s, the internet has largely used loss-based congestion control. In October ’86 a series of congestion collapses started which led to new investigations/researches on the matter, and the implementation of the above mentioned type of congestion control. Though this was a good solution at the time – and worked well for many years since, in today’s complicated networking world it has become rather problematic. Enter Google’s BBR, an ambitious solution that rewrites congestion control entirely.
BBR looks like Google’s next big step in their long-term goal to make everyone’s internet faster. After TCP changes, Chrome and the QUIC protocol, there was a need for an algorithm that can respond to actual congestion, not just packet loss. This is where BBR comes in. Google’s engineers started from scratch, with an entire new paradigm: to decide how fast to send data over the network, BBR considers how fast the network is delivering data, therefore the network connection is an important variable. BBR uses recent measurements of the network’s delivery rate and round-trip time in order to build an explicit model containing both the maximum recent bandwidth available to that connection, and its minimum recent round-trip delay. This way, the algorithm can control how fast it sends data and the maximum amount of data it’s willing to allow in the network at any time – a great solution for when data routes become congested.
How do Google Cloud customers benefit from it?
– Higher throughput, lower latency and better quality of experience across Google services (compared to CUBIC).
According to Google’s blog post, there are 2 main ways that Google Cloud Platform customers can benefit from BBR: from GCP services to cloud users, and from Google Cloud to internet users.
Cloud Spanner, Cloud Bigtable or Cloud Storage send the traffic from the GCP service to the application using BBR, which results in faster access to data for cloud users.
Google Cloud Load Balancing or Google Cloud CDN can be used by GPC customers to serve and load balance traffic for their website. The users of your website get faster downloads because (you guessed it) the content is sent to users’ browsers using BBR.
“BBR allows the 500,000 WordPress sites on our digital experience platform to load at lightning speed. According to Google’s tests, BBR’s throughput can reach as much as 2,700x higher than today’s best loss-based congestion control; queueing delays can be 25x lower. Network innovations like BBR are just one of the many reasons we partner with GCP,” said Jason Cohen, Founder and CTO, WP Engine.
YouTube’s 4 percent higher network throughput was achieved with BBR because the algorithm discovers the most effective way to utilize the bandwidth offered by the network. Google’s blog posts explains that BBR also keeps network queues shorter, reducing round-trip time by 33 percent; this means faster responses and lower delays for latency-sensitive applications like web browsing, chat and gaming. Moreover, by not overreacting to packet loss, BBR provides 11 percent higher mean-time-between-rebuffers. This might not seem like much on a first look, but on a large scale these improvements become remarkable.
The Google Cloud Platform is evolving constantly, there’s even a public bbr-dev e-mail group available for those interested in participating in discussions, keeping up with the latest BBR news or even pitching in on open source BBR development or testing.
All the benefits describe above come for users of Google’s Cloud Platform with no additional cost. Using technologies like Espresso, Jupiter, Andromeda, gRPC, Maglev, Cloud Bigtable and Spanner will activate your BBR-improved internet speed.
