Subscribe and receive the latest news from the industry.
Join 62,000+ members. Yes it's completely free.
Google is planning to use the same congestion control algorithm it uses with search and YouTube to give the company's cloud platform a speed boost.
July 20, 2017
Google is looking to give its public cloud platform the same speed boost that powers the company's search engine, as well as YouTube.
On Thursday, Neal Cardwell, a senior staff software engineer, published a blog post, which describes the company's efforts to bring its congestion control algorithm TCP BBR to the Google Cloud Platform.
BBR stands for Bottleneck Bandwidth and Round-trip propagation time, which Google developed in-house and started using with www.google.com and YouTube in 2016. These sorts of congestion control algorithms run inside PCs, smartphone and other devices that are connected to a network and decide how fast to send data from one point to another.
So what's the difference between BBR and other algorithms?
Since the early 1980s, when the roots of the Internet were taking shape, the majority of TCP (Transmission Control Protocol) congestion control algorithms -- Google used CUBIC for many years -- relied on packet loss to determine whether there was a congested connection between two points. This worked effectively early on as these algorithms were able to match up well with low-bandwidth Internet links.
Figure 1: (Source: DWilliam via Pixabay)
However, as networks got bigger, and the Internet much faster, Google engineers decided to look at these algorithms and create a new technology that would address bottlenecks in the network before they happened. This is where BBR comes in.
As Caldwell writes in the July 20 post:
We need an algorithm that responds to actual congestion, rather than packet loss. BBR tackles this with a ground-up rewrite of congestion control. We started from scratch, using a completely new paradigm: to decide how fast to send data over the network, BBR considers how fast the network is delivering data. For a given network connection, it uses recent measurements of the network's delivery rate and round-trip time to build an explicit model that includes both the maximum recent bandwidth available to that connection, and its minimum recent round-trip delay. BBR then uses this model to control both how fast it sends data and the maximum amount of data it is willing to allow in the network at any time.
When Google applied BBR to YouTube, the company claims it led to a 4% higher throughput on the site, as well as lower latency.
Keep up with the latest enterprise cloud news and insights. Sign up for the weekly Enterprise Cloud News newsletter.
Now, Google is applying BBR to its GCP offering, which should make it faster to retrieve data from the cloud. This is especially important as data is moving, for example, from Europe to the US, or vice versa, and the users of a cloud service want as little lag time as possible, especially if the customer is crunching a big number set in database.
Google is also promoting the BBR technology as a way to serve and load balance traffic going to websites. This could help increase download speeds for users.
"The end result is faster traffic on today's high-speed backbones, and significantly increased bandwidth and reduced download times for web pages, videos, or other data," Caldwell writes.
Managing Editor, Light Reading
Prior to joining Enterprise Cloud News, he was director of audience development for InformationWeek, where he oversaw the publications' newsletters, editorial content, email and content marketing initiatives. Before that, he served as editor-in-chief of eWEEK, overseeing both the website and the print edition of the magazine. For more than a decade, Scott has covered the IT enterprise industry with a focus on cloud computing, datacenter technologies, virtualization, IoT and microprocessors, as well as PCs and mobile. Before covering tech, he was a staff writer at the Asbury Park Press and the Herald News, both located in New Jersey. Scott has degrees in journalism and history from William Paterson University, and is based in Greater New York.
You May Also Like
Rethinking AIOPs — It's All About the DataMar 12, 2024
SCTE® LiveLearning for Professionals Webinar™ Series: Fiddling with Fixed WirelessMar 21, 2024
SCTE® LiveLearning for Professionals Webinar™ Series: Cable and 5G: The Odd Couple?Apr 18, 2024
SCTE® LiveLearning for Professionals Webinar™ Series: Delivering the DAA DifferenceMay 16, 2024