Google Challenges Amazon With Compute Engine
Google today announced the long-awaited expansion of its cloud computing services to compete directly with Amazon Web Services. Google Compute Engine is an infrastructure-as-a-service (IaaS) offering.
June 28, 2012
google-servers-1
Cloud computing users can now run their applications on virtual machines in Google's data centers, providing a scalable IaaS service for Linux apps (Photo: Google)
Google today announced the long-awaited expansion of its cloud computing services to compete directly with Amazon Web Services and its industry-leading suite of cloud services. At its Google I/O conference, the company rolled out Google Compute Engine, an infrastructure-as-a-service (IaaS) offering that allows users to run applications in virtual machines hosted in Google's data centers. The new compute service will combine with existing offerings - including Google App Engine, Google Cloud Storage and Google BigQuery - to form the Google Cloud Platform.
Compute Engine was developed "in response to many requests from developers and businesses," Product Manager Craig McLuckie wrote on the Google Enterprise blog. McLuckie said the service will launch in limited preview, but "our goal is to give you all the pieces you need to build anything you want in the cloud. We’re just at the start of what the cloud can do"
But not at the start of the public cloud. Amazon Web Services launches its Elastic Compute Cloud (EC2) service in August 2006, meaning it has nearly a six-year head start on Google in IaaS services. By some estimates, Amazon EC2 may be running as many as 450,000 servers.
Google Compute Engine can launch Linux VMs on-demand. VMs are available with 1, 2, 4 and 8 virtual core with 3.75GB RAM per virtual core. Virtual machines can be managed via a scriptable command line tool, web UI or user-created management system using Google's API. The service will launch with support from many leading players in cloud management, including RightScale, Puppet Labs,OpsCode, Numerate, Cliqr and MapR.
"In order to help our customers get the most out of our cloud platform products, we have worked closely with technology partners to integrate complementary offerings and with services firms to enable them to build powerful new cloud-based solutions that help customers accelerate their success and innovation." explains Shailesh Rao, Director of New Products and Solutions for Google Enterprise.
"Google has certainly already proven itself in the cloud space and Google Compute Engine will be a major player in the IaaS arena," said Thorsten von Eicken, CTO of RightScale. "It was clear from day one that Google Compute Engine is an all-out initiative drawing on an incredible pool of expertise and technology to deliver a world-class product. This is an important addition to the Google portfolio of cloud services -- allowing customers to have freedom to use PaaS and IaaS or both, as appropriate for their use case. We're pleased to be a launch partner and to have customers already running workloads on Google Compute Engine."
BuildFax, a RightScale customer, described its early experience with the Compute Engine service.
"We have been using the Google Compute Engine for our web application (nginx, memcached and rabbit) and database (MySQL) servers, and have found the performance of the Google Compute Engine VMs to be the most consistent of any other virtualized architecture we've used," said Joe Masters Emison, VP, Research and Development at BuildFax. "BuildFax creates millions of construction history reports each month, and we have extensive benchmark tests that evaluate how any particular hardware platform will perform in delivering reports to our customers. Using Google Compute Engine 8-CPU instances, each report took an average of .39 seconds, with a standard deviation of .02 seconds; we are used to seeing a standard deviation of almost ten times that on other virtualized hardware solutions. The only way we've found to get that kind of consistent performance -- before Google Compute Engine - is running our own data center."
About the Author
You May Also Like