AWS vs. Google comparison guide
A comprehensive collection of articles, videos and more, hand-picked by our editors
Google Compute Engine will soon include important features that are sure to put the IaaS platform on a more even playing field with Amazon Web Services, according to industry sources.
Two of the key features Google Inc. must add to its Google Compute Engine (GCE) to be on par with Amazon Web Services (AWS) are autoscaling, a cloud computing service feature that automatically adds or removes compute resources depending upon actual usage, and reserved instances, which allow customers to purchase capacity for a fixed period of time for a lower fee than on-demand instances.
Sources close to the company say these two features are on their way.
It's unclear exactly when the new features will launch, but there is a Google Cloud Platform event scheduled for March 25, which will include an "overview of a brand new cloud platform service," according to the event website.
"Autoscaling is huge, and reserved instances are huge too," said one large enterprise customer, speaking on condition of anonymity, who said Google shared a product roadmap that includes both new features. "They are musts for running with large groups of users in the cloud."
Kyle Hilgendorfanalyst, Gartner Inc.
An IT services provider and Google partner based in Chicago echoes the desire for these features, particularly reserved instances.
"One of the challenges with Google has been lack of reserved instances," he said. "If we connect Salesforce to Google Search appliance, it's always running, so our option has been to either go with Google and pay a lot or use AWS Reserved Instances, which is cheaper."
Google IaaS vs. AWS
Others not privy to Google's roadmap said both of these features are must-adds for Google as it looks to compete with AWS in Infrastructure as a Service (IaaS). AWS offers a service that allows customers to create Auto Scaling Groups among Elastic Compute Cloud instances.
"They're going to need to start doing what [Microsoft] Azure did," in terms of playing catch-up with AWS, said Kyle Hilgendorf, an analyst with Gartner Inc., a research firm based in Stamford, Conn. He has not been briefed on Google's plans but said autoscaling and reserved instances should be priorities for the company as it branches out in IaaS.
"Right now the feature gap [between GCE and AWS] is too much for most enterprises," Hilgendorf said. "But if a year from now they can match AWS feature-for-feature, they will garner more interest."
Performance is where Google IaaS claims an advantage over AWS already, as it offers a private backbone network which it says causes less contention for resources among virtual machines.
The Chicago-based IT services provider said using Google is preferential to AWS in some ways, but there are advantages to using Amazon's cloud.
"The infrastructure is more mature in Amazon, but the performance reports show that Google's infrastructure performs better, and you have no fee for bringing data into the platform," he said. "However, the advantage of AWS is that many companies are already on it."
Google Compute Engine launched as a tech preview in the summer of 2012; it followed another service that makes up the Google Cloud Platform called Google App Engine, a Platform as a Service offering that launched in 2008 based on a programming platform first used internally by Google engineers. GCE was made generally available in December, with IaaS prices competitive to AWS.
Google App Engine has offered a framework for developers to build their own autoscaling, but it is not yet a native feature of GCE.
Google declined to comment.
Senior News Director Bridget Botelho contributed to this report.