Amazon Web Services (AWS) comprises more than 30 services in dozens of data centers located in nine regions around the world. The task of fully comprehending exactly what AWS is and how it can help your business can be daunting, but it doesn’t take long to realize that the depth and breadth of AWS is significant.
AWS offers deployment, management, computing, storage, networking, as well as a host of supporting services such as queues and email services. Chances are AWS has a service or two to help your company work faster, smarter, and more cost effectively.
I want to start by providing a good understanding of AWS, how it works, and how your company can get started benefiting from its services.
A Quick History Lesson
Like many surviving startups of the dot-com era, Amazon found itself with an enviable problem at the turn of the century. The sheer scale of their business had grown beyond the capacities of any available pre-packaged software solutions.
Amazon had to completely re-think their vast infrastructure from the ground up. Having to re-design all their systems to deal with a new set of requirements that their users had, Amazon set out to ensure this new infrastructure would provide high availability, auto scaling, ease of use, as well as integrated backup and disaster recovery.
Amazon also knew their new infrastructure needed infinite scale. Users should not have to consider ever “outgrowing” a service. Their Simple Storage Service (S3) allows users to store an unlimited number of objects while their Elastic Compute Cloud (EC2) lets users spin up thousands of virtual servers.
When Amazon finished building much of their required infrastructure in the early 2000s, they realized they had another enviable problem: extra capacity that they rarely and sporadically used. In 2006, they opened AWS for limited public beta, and although the offerings were only a fraction of what is currently available, the product line became wildly popular. By 2007, they had attracted more than 300,000 users, and in 2008, the “beta” moniker was dropped.
Subsequent years have seen rapid releases of new services, features added to existing services, and prominent customers (including Pinterest, NASA, and Netflix).
In addition to the previous requirements, Amazon tacked on new features to serve external customers, such as flexible pricing, heightened security, and proactive price cuts.
Proactive Price Cuts
Because AWS operates with significant scale, they can purchase hardware, software, power, bandwidth, and nearly everything else at much, much lower prices than competitors.
Instead of pocketing those savings as profits, AWS has traditionally passed those savings on to customers, reducing prices over 30 times in the last six years.
For these reasons and many more, AWS has become an attractive way for businesses of all sizes to deploy and serve the computing and storage needs of their customers. For some companies, such as Netflix, trusting AWS with all of their computing needs enables them to operate with significantly reduced headcount and a great deal more agility.
AWS calls their individual data centers “Availability Zones” or “AZs.” Each data center has multiple redundant power and bandwidth providers, and AZs are organized into “Regions”—collections of physically close data centers that are interconnected via high-bandwidth, low-latency fiber. Amazon currently has nine Regions (although one is strictly for US government use), and each Region has between two and five AZs.
Most customers will decide to deploy in one Region and use multiple AZs within that region for high availability and disaster recovery. Some services, such as the storage product Amazon S3, deploy to a specific Region, using multiple servers at multiple AZs to allow stored objects to be both incredibly redundant and fault tolerant. Other services, like the virtual server product Amazon EC2, deploy to only one AZ (you can only launch a virtual server into a single data center). However, by using Auto Scaling groups, one could launch a few servers in two or more AZs and then balance across multiple AZs within the region to achieve redundancy and fault tolerance.
These “multi-AZ” deployments are the preferred way to serve web and mobile traffic within AWS.
In addition to Regions and AZs, AWS also provides dozens of lightweight caching servers known as “Edge Locations.”
Currently used only by the content delivery network (CDN) product called Amazon CloudFront and the DNS product Amazon Route 53, these Edge Locations cache content and DNS records closer to individual users, whereby reducing latency and increasing web page load times. AWS currently has 42 Edge Locations around the globe.
AWS is aggressively adding AZs, Regions, and Edge Locations to serve users better. End customers get immediate access to new locations as they are added.
This content was adapted from the Global Knowledge white paper Amazon Web Services: An Overview.