AH Google Logo  - Chris 2015 -31

Google’s ‘Seesaw’ Load Balancer Goes Open Source

January 29, 2016 - Written By Daniel Fuller

If you’re a network or systems administrator, you’re likely familiar with the concept of a load balancer. It’s a hardware device or software stack that distributes network application load across all the machines and servers connected to it in order to help mitigate network congestion. Google’s software solution, called Seesaw, was created in 2012 in response to a lack of adequate load balancing software for Google’s own use. Coded in Google’s own Go language, the software boasted a flexible Linux backbone and was used to manage Google’s own network needs, which entailed things like automated deployment and ease of use and maintenance.

The system has served Google’s needs quite well for some time now, but they have so far not leased the software out to anybody, likely for support reasons. The software was custom made for Google’s hardware, software and network backend, so leasing it out to another company would entail helping to shape the solution to the company’s own specifications or reshaping the company’s entire backend to match Google’s. This very approach is a big reason some analysts think that Google’s Cloud Infrastructure As A Service isn’t doing as well as it could be doing. That being the case, making the software open source was apparently the most logical choice. A new, open-source variant of Google’s in-house Seesaw load balancer is now available on Github under the fairly lenient Apache open source license.

This release is not meant to be a full endorsement of support from Google, of course – users and companies who compile and use it cannot expect any support from Google, nor can they expect updates from the official product. Although upstream updates are a possibility, support is likely much less so, since that would require an investment of time and money from Google. No word on a yes or no for updates or support was given in the blog post. Under the Apache license, given the proper credits and pointing out of changes, anybody is free to modify and use the code as they see fit, so long as they don’t claim the software as their own and sell it. Hit up the source link for all of the details on the official blog post, as well as the Github link.