One of Amazon Web Services’ most potent weapons in the cloud-computing wars with Microsoft and Google is also one of its least understood.
About a year ago, Amazon introduced Lambda, which it describes as “a compute service that runs your code in response to events and automatically manages the underlying compute resources for you.”
That sounds like a lot of buzzwords, but developers and customers like Major League Baseball and Twitter’s Periscope love Lambda because of how it makes running apps simpler. So far, none of Amazon’s competitors has come up with anything similar.
“As a technologist, there’s almost always a ‘gotcha,'” says Matt Wood, Amazon Web Services’ general manager of product strategy. “Lambda kind of is as magical as it appears.”
And it’s saving some of Amazon’s customers big bucks, with at least one happy Lambda customer saving 80% off their cloud bills.
It’s just another way that Amazon is extending its lead in the hot cloud-computing market, where it’s already a $7 billion player and growing fast.
When Amazon Web Services launched in 2006, its first feature was the Simple Storage Service, or Amazon S3, which made it simpler for programmers to store things like files on the web. You just sign up for an account and you have a place to stick files for use in websites.
“[Lambda is] the same idea as S3, but applied to computation,” Wood says.
Building an app used to require a lot of infrastructure. If you’re building a video-streaming app, for instance, you need to set up a whole bunch of servers to upload videos, transcode it into the format you want, and play it at different resolutions and sizes depending on the kind of device the customer is using.
But it also means that you had to build big, complex systems that might sit idle 99% of the time, except for the 1% of the time when you have a new video coming in.
This is what makes AWS Lambda so special. With a snippet of specialized code, called a “Lambda function,” an app can “talk” to the Amazon Web Services cloud, access the functions it needs, and then turn it off afterward — without the programmer having to worry about a single server.
[to continue, click HERE]