Open Source and Cloud Computing

Yahoo has open sourced a version of its Traffic Server, a high-performance application server for building cloud services that it got when it acquired Inktomi. It donated the code to the Apache Software Foundation through the Apache Incubator.

It says Traffic Server enables the session management, authentication, configuration management, load balancing and routing of an entire cloud computing stack.

It’s supposed to offer fast, reliable and scalable access to cached online content and speed responses to requests for stored web objects, such as files, news articles or images, reducing bandwidth usage and costs.

Yahoo says its low-latency, extensible framework makes it ideal for delivering web traffic at high rates, and its “plug-in” architecture makes it customizable to fit different system needs.

The company previously open sourced its version of Hadoop, the Google MapReduce-inspired software framework and file system for distributing and running applications on clusters of servers.

Having open sourced two pieces of code, the company is preening about its “unprecedented” commitment to open source cloud computing initiatives. And it says other cloud technologies will follow as they mature.

Shelton Shugar, Yahoo’s senior vice-president of cloud computing, calls Traffic Server an “essential building block for cloud computing,” and integral to Yahoo’s own edge services, on-line storage and cloud serving.

It reportedly represents more than eight years of active use and engineering and currently serves more than 30 billion web objects a day across the Yahoo network.

Yahoo’s global network of data centers let Traffic Server choose the closest servers to store and access cached content for increased speed. It can handle more than 30,000 requests a second per server and currently serves more than 400 terabytes of data a day.

Yahoo means to build a community of developers around the open source Traffic Server to take it to the next level.

Yahoo also announced an update to Hadoop.