The Autonomic Computing Journey: The Online Bookstore That Changed the World
So far in this series, we've been focusing on history. Moving onto more recent events, see how Amazon fundamentally altered computing and the cloud.
Join the DZone community and get the full member experience.Join For Free
Following along our autonomic computing journey so far (Part 1, Part 2, Part 3), we have caught up to the more recent generations of IT. Something happened as the IT environments evolved in the decade that led up to the post-Y2K era. No matter where you are in your journey through this history, I would bet you know this part of the story well.
You’ve inevitably been at a tech conference, and we joke about how Infrastructure-as-a-Service is a challenge to deploy, and that competitors find themselves up against a “bookstore.” Obviously, the Amazon story has become one that shaped the industry. Bookstore or not, Amazon Web Services did something that no virtualization vendor could: it changed the world of IT consumption.
Agility Inside, Sellable Platform Outside
The services that we know now as the AWS platform were once internal services. The Elastic Compute Cloud (EC2) and the Simple Queue Service (SQS) were internal services which powered the Amazon retail environment. In order to scale at incredible rates, and to respond to bursts, the engineers at Amazon developed their own internal practices, and the architecture on top of which they would develop and scale.
Because the requirements for companies like Amazon are unique, it drives them to engineer their own solutions. At the time, there were no public clouds that could be consumed and scaled to meet their needs. There was no AWS at that time, yet. This was the dawn of the first significant private cloud architecture.
Private Becomes Public
Jeff Bezos recognized the value of the AWS internal architecture. This value transcended their own requirements and became an obvious fit to others in the industry. From this was born the first iteration of AWS, which launched as a publicly consumable cloud in 2004 as they opened the doors to SQS as an available system to all. EC2 followed shortly afterward, and the environment was touted as having 180,000 developers within three years ,according to some.
We know how the story has gone from there. There has been some significant growth in services and regions since those early years, as the current AWS console will tell you.
If there is any doubt about the value that has been gotten from this venture, you can look at the freshly announced market capitalization which places AWS above Exxon. Amazon is now the fourth largest publicly traded company in the world.
Not bad for a bookstore, right?
The reason we talk about AWS on the journey towards autonomic computing is the effect that they have created. The idea that IT can be consumed easily, and on-demand, was a fundamental shift. It created the true market for public cloud infrastructure. This is also what triggered the next phase in enterprise computing: private cloud.
Many organizations sought out the same service-oriented IT architectures inside the walls of their on-premises data centers. The issue with creating private cloud architectures was the lack of understanding of how to do it at scale, for a large and diverse audience of consumers. Amazon had the distinct advantage of building services that they needed, which could simply be opened up to public consumption. You get what you get, and Amazon builds what it feels like.
Early financial reporting didn’t even split out the cloud side of the balance sheet, which gave Amazon the advantage of keeping their sales in that sector quiet so as not to expose potential shortfalls on how their services were truly being consumed by outside customers.
Services would come to be the panacea of the developer. API-accessible infrastructure, available on-demand, and in a composable, recipe-driven approach. What if we wanted the same consumption model, but didn’t necessarily need the rapid elasticity? What if the boundaries of the physical data center were real due to regulatory challenges around data sovereignty?
Cloud as a concept was winning over many developers. It was also dripping off the tips of the tongues of many CIOs. This is where it became interesting, because we were about to find out just how challenging the private cloud world was. That is the next step in our journey.
Published at DZone with permission of Eric Wright, DZone MVB. See the original article here.
Opinions expressed by DZone contributors are their own.