New features and tutorials for Compute Engine Load Balancing

Modern web apps, such as online banking apps, often need to keep session information, which requires special support from the load balancing so that you keep getting connected to a server that has the necessary context information to help you. With the latest update to Google Compute Engine Load Balancing, you now can spread the load while tracking sessions using the client IP address. When enabled, this feature ensures that all requests from the same IP (in a session) get forwarded to the same healthy Compute Engine instance.



In addition, we’re introducing backup pool, which can add more redundancy to your load balanced website or application. When the health of the servers in the primary pool drops to some configurable threshold, the servers in the backup pool will take over until the primary pool is healthy again. This feature can also give you more flexibility and convenience especially during patching and upgrading of your backend instances.



Both features are supported via the gcutil command line tool, RESTful API and the Cloud Console. Please read the documentation and start using load balancing today. Also check out the following tutorial demonstrating how quickly you can set up load balancing.







If you want to learn more, we’ve published an in-depth look at load balancing on Compute Engine that shows how easy it can be to set up for even complex apps. In addition to discussing use case scenarios, this article presents the results of running load tests with a half million requests to illustrate how our load balancing works. If you've been wondering whether load balancing is right for you, read the article to find out (spoiler alert: if your application scales to more than one Compute Engine instance, the answer is yes).

Web application using Compute Engine Load Balancing

Happy load balancing!



-Posted by Gary Ling, Product Manager