For the past few weeks I have been keeping a secret. It's the first big secret project (with code names, close the blinds when reviewing the deck, etc) I have ever worked on and it's been a little annoying because I really wanted to talk about this.
As far as secret projects go, this one was actually pretty public. Several bloggers came pretty close to describing what Ray Ozzie announced at PDC today. I should also disclose that my involvement has been limited to this point, I have effectively had 8 days in the group and a good chunk of that was spent trying to figure out what it is I do here.
When I was going through the interview process I had a vague idea of what we were going to announce but was really sold by opportunity to launch an incubation project at Microsoft into the market. I was told that almost 200 engineers, some of the best at the company, had worked on the product for nearly two years. And that it would change the industry...
Introducing Windows Azure
As you may have seen today Microsoft announced Windows Azure, which is an operating system for the Cloud along with versions of SQL Server and .NET framework optimized for this Cloud platform. It will enable developers to create Web based applications and not worry about servers as they will run in Microsoft data centers. Applications will be scalable on demand with the ability to make real-time updates with no disruption of the service. It will leverage tools and frameworks developers already use. The investment to bring this to market is staggering, the decision to use the Windows brand, arguably Microsoft's most valuable asset, is an indication of how serious the company is about this new offer.
How Windows Azure Helps Brings Products to Market Faster
For the non-technical people you'll have to take my word that this is really cool and a big deal. Here's an example of why. A few years ago I was an IT manager at a mid-size company before I realized it was a mistake and I was supposed to be back in marketing. My team built a Web based configurator that let people design their high school and college class rings online. This software was used on our own 20+ Web sites as well as on some of the largest e-commerce sites at the time like Wal-Mart.com. People only buy class rings over a two month period near the beginning and end of the school year. But the site we built had to be ready to support those big spikes in demand basically all the time. That largely impacted decisions about the server infrastructure.
If you think about your computer at home, when you try to do too much with it at once things run really slowly. Web servers are the same. At some point when too many people are trying to use them they slow down or crash and customers tend to get frustrated by that and go somewhere else. Because Wal-mart.com and the other retailers required us to have the ability to handle those spikes in demand basically year round, the servers were usually running at a fraction of their capacity the vast majority of the time. This was wasted investment and a lot of money for our smaller company.
The Windows Azure platform would have made a big difference for this type of project. Instead of having to procure those servers that sat idle most of the time we could have spent a fraction of the money and hosted the application in Microsoft data centers. We would have paid only for what we used. There was also a very large amount of time that my team and I spent on buying equipment, hosting equipment, patching equipment, monitoring equipment, testing equipment, giving the equipment funny names, and generally worrying about--the equipment.
All that time and effort was not helping us build a better product.
If this platform had existed at the time, our project would have been live much faster, would have cost much less and probably be a richer experience because of the time we could have reinvested from thinking about servers back into the software.
Why Cloud Computing Is Changing IT
Cloud computing is the next big iteration of the IT industry. Its impact will be similar to the change from mainframe to client-server. Some of the change will be fast and visible, much of it will not. It's more likely that we will look back at some point and realize how big and disruptive this trend turned out to be on the industry. Cloud computing has grown from a theory to an experiment to increasingly mainstream over several years. Much of what was required to make it possible has only come to fruition in the last few years.
In the not so distant past, companies and services that operated purely online were a novelty. But increasingly people became comfortable with the idea of having their email exist only on the Internet and not inside their own computer (Hotmail, Gmail, etc). Today it is considered table stakes for banks to have online portals to view account details, but there was a time when people were really freaked out about having their financial information on the Internet.
As we began to accept the Cloud in our personal lives we brought that acceptance to work with us. Soon we thought it was not only acceptable but really convenient to have our customer information and sales plans online (SalesForce.com) and we were also ok with the idea that all that information would not only be online only but held exclusively on another company's data center. As we traded the perceived (but rarely realized) dangers of having our personal and work information online the infrastructure that made this possible became increasingly affordable.
What It Is I Do Here
For several years now my friends and family (and sometimes my co-workers) have not really understood what it is I actually do at work. Unfortunately that trend will probably continue with this role as well. Basically my job is figure out how to take Windows Azure to market after the preview stage we are in today. It is a very exciting challenge, fascinating technology, and the opportunity to work with brilliant people. It is exciting to finally be able to talk about this.
