Another open-source consortium for Facebook, new project to go along
Even though it doesn’t depend on developers for revenue as much as some of the other web-scale companies that have been making headlines in the ecosystem lately, Facebook Inc. is investing heavily in in open-source. At its @Scale conference in San Francisco this week, the social networking powerhouse unveiled a new initiative aimed at bringing together the industry’s most prominent Internet firms to drive the adoption of community-led projects in the traditional enterprise.
The program joins the Open Compute Project, an effort Facebook launched in 2011 to share the designs for the specialized hardware powering the world’s largest and most efficient data centers, including its own, with the broader industry. The goal of the initiative is to make those technologies not only available for traditional organizations but also practical to implement, a vision that new initiative extends up the stack to software.
According to Facebook open-source lead James Pearce, the TODO consortium – whose name is a clever acronym for its mission statement of “talk openly, develop openly” – will focus on spreading the lessons that the web-scale crowd has learned in the course of its interactions with community-led projects. The social networking giant has a particularly rich history in that arena, having developed the Cassandra document database, the HipHop source code accelerator for PHP and several other key open-source technologies.
The latest addition to the bunch is something called Mcrouter, a protocol released in conjunction with the debut of TODO at @Scale designed to automate the management of disparate cache pools that employ the popular memcached key-value store. Written in C++ and C, the technology provides a centralized routing layer that monitors the health of every node, automatically redistributes requests in the event a pool becomes unresponsive and reverts to the original arrangement as soon as it comes back online.
Mcrouter also makes it possible to send the same request to multiple destinations, replicate data across multiple servers to reduce the risk of downtime and perform a host of other operations that had up until now required users to jump through hoops to pull off. The code for the tool is available under a Berkeley Software Distribution license.
Facebook divulged that it has implemented Mcrouter in thousands of servers across its network of data centers, with the technology now serving as many as five billion requests per second at peak times. But firm is still keeping a tight lid on exactly what it is that TODO is doing, although the official website for the initial does offer some details. The project apparently includes 9 other companies besides Facebook, including Twitter Inc., Google Inc. and a number of other familiar names. Pearce promised in his blog that more information will be released on the effort in the coming weeks.
photo credit: mkhmarketing via photopin cc
A message from John Furrier, co-founder of SiliconANGLE:
Your vote of support is important to us and it helps us keep the content FREE.
One click below supports our mission to provide free, deep, and relevant content.
Join our community on YouTube
Join the community that includes more than 15,000 #CubeAlumni experts, including Amazon.com CEO Andy Jassy, Dell Technologies founder and CEO Michael Dell, Intel CEO Pat Gelsinger, and many more luminaries and experts.
THANK YOU