Autoscaling, welcome to Google Compute Engine
Posted:
Monday, November 17, 2014
The true power of cloud computing is unlocked when developers can build resilient and cost efficient applications that use just the right amount of resources necessary at any given time. So the same team that designed the scaling infrastructure for products like Google Search and Gmail have brought a highly anticipated feature to Google Compute Engine - intelligent horizontal Autoscaling. Today we are releasing the service into Beta, which means it is now available for everyone to start using.
Autoscaling allows customers to build more cost effective and resilient applications. Using Compute Engine Autoscaling, you can ensure that exactly the right number of Compute Engine instances are available at any given time to handle your application’s workload. This saves you money when your application’s usage is low, and ensures your application is responsive when utilization is high.
The Compute Engine Autoscaler is able to intelligently and dynamically scale the number of instances in response to different load conditions by defining the ideal utilization level of a group of Compute Engine instances. This means that when the actual utilization of your service increases or decreases, Autoscaler will detect the change and adjust the number of running instances to match. Autoscaler can respond to a number of different metrics such as CPU load, QPS on a HTTP Load Balancer and metrics defined using the Cloud Monitoring service.
One early customer of Compute Engine’s Autoscaler was Wix.com, the popular website-building service. Golan Parashi, Wix.com's Infrastructure Team Lead, commented how Google uses heuristics to determine how many instances to add at one time to hit demand, “reducing [our] expenses, while giving us confidence that Google will manage the appropriate number of machines, even when a spike occurs."
Autoscaler not only chooses the right number of instances but also adapts automatically based on how far the current state is from the desired target. This means Autoscaler performs well even in unexpected scenarios such as sudden traffic spikes. At Google Cloud Platform Live, we demonstrated how an application could scale from zero to handling over 1.5 million requests per second using Autoscaler.
Here are some additional resources to get you up to speed on Compute Engine’s Autoscaler:
We can’t wait to see what you build - and scale - next on our platform.
-Posted by Filip Balejko, Software Engineer
Autoscaling allows customers to build more cost effective and resilient applications. Using Compute Engine Autoscaling, you can ensure that exactly the right number of Compute Engine instances are available at any given time to handle your application’s workload. This saves you money when your application’s usage is low, and ensures your application is responsive when utilization is high.
The Compute Engine Autoscaler is able to intelligently and dynamically scale the number of instances in response to different load conditions by defining the ideal utilization level of a group of Compute Engine instances. This means that when the actual utilization of your service increases or decreases, Autoscaler will detect the change and adjust the number of running instances to match. Autoscaler can respond to a number of different metrics such as CPU load, QPS on a HTTP Load Balancer and metrics defined using the Cloud Monitoring service.
One early customer of Compute Engine’s Autoscaler was Wix.com, the popular website-building service. Golan Parashi, Wix.com's Infrastructure Team Lead, commented how Google uses heuristics to determine how many instances to add at one time to hit demand, “reducing [our] expenses, while giving us confidence that Google will manage the appropriate number of machines, even when a spike occurs."
Autoscaler not only chooses the right number of instances but also adapts automatically based on how far the current state is from the desired target. This means Autoscaler performs well even in unexpected scenarios such as sudden traffic spikes. At Google Cloud Platform Live, we demonstrated how an application could scale from zero to handling over 1.5 million requests per second using Autoscaler.
Here are some additional resources to get you up to speed on Compute Engine’s Autoscaler:
- Watch us automatically scale up to 1 million queries per second while on stage talking about Autoscaler at Google Cloud Platform Live
- Learn more about Autoscaling on Google Compute Engine
- Learn more about HTTP Load Balancing
We can’t wait to see what you build - and scale - next on our platform.
-Posted by Filip Balejko, Software Engineer