Google v Amazon: It’s Not Just A Price War

Google has been getting a load of press attention about the pricing challenge it has leveled against the cloud services industry, especially versus the world’s largest cloud services provider, Amazon Web Services, and Microsoft.

When asked if Google and Amazon were in effect waging a price war, Google communications director Danielle Aronstam told eWEEK that “the announcements we made are not in response to pricing updates that are competitive. We really look at the concept of Moore’s law. We’re basing pricing on the costs that we’re seeing, and what costs that we should be passing through to our customers.”

Cutting prices in the cloud

AWS immediately announced its 42nd price reduction since it started in the cloud services business in 2006. Google had announced several price cuts – ranging from 32 percent to 85 percent – in some of its services.

Pricing aside, the  announcement from Google was also about the near-completion of the Google Cloud Platform, which has been put together piece by piece during the last four or five years. We say “near-completion” because platforms such as these are never completed; they’re constantly morphing into new and improved versions. At least that’s the story cloud service providers tell, and they’re sticking to it.

The cloud-building platform  now includes developer-friendly features such as a new Google APIs Client Library for .NET and improved documentation for using third-party Puppet, Chef, Salt and Ansible configuration-management tools.

Google offered evidence as to how it is building itself into a more instrumental player in cloud services for developers. It is bringing much of the work done inside of Google for internal use to the outside in order to improve developers’ productivity.

No trade-offs for developers

Greg DeMichillie, director of product management for Google Cloud Platform, told attendees at an invitation-only launch event in San Francisco that all the updates are designed to enable developers to easily build applications without having to worry about trade-offs — most of which involve scale versus performance.

Perhaps the most important news of the day was that Google Compute Engine will soon be offering new virtual machine-making options that include three industry mainstays: Red Hat Enterprise, SUSE Enterprise and Microsoft Windows Server 2008. The expanded offerings not only open a world of new options for developers, but they also make it a more attractive competitor to Amazon Web Services, which has ruled the roost since 2006.

In the VM world, Google still has a way to go before obtaining parity with AWS, which already offers a slew of virtual machine-making options — including several older versions of Windows Server.

Here are the highlights added to Google Cloud Platform:

• An improved BigQuery data streaming utility, which now has the ability to ingest data 100 times faster than previously. Where it used to ingest 1,000 rows of data per second, it now can do up to 100,000 rows per second. DeMichillie offered the use-case example of the Glen Canyon power company in Seattle, which processes 17 billion events (8,000 to 10,000 per second) and 2TB of data per month in monitoring records from 400,000 power meters in Seattle, and then can quickly process it all to gain instant insights.
• The Managed VM service in the Google App Engine application hosting platform will now allow developers to build Web services as usual, but then be able to use a virtual machine to handle custom functions that App Engine doesn’t support. Thus, users can use the auto-management of App Engine and any virtual machine they want.
• Updates to the developer dashboard make it easier to manage code in the cloud. For example, when a developer commits some new code, the system will now automatically compile it and run specified unit tests on it. The platform now makes it easier for developers to track down bugs by providing one-click access to problematic code and detailed stack traces.
• A new ‘”Diff” link to the Releases page will bring users to a rollup of all the commits that happened between deployments and the source code changes they introduced. “This is invaluable when you are trying to isolate a production issue,” Product Manager Cody Bratt wrote in the Google Cloud Platform blog.

Originally published on eWeek.