# Monday, November 3, 2008

I have been a fan of the cloud since Amazon first released its first APIs. We have been waiting for Microsoft to enter in the cloud space and we have been seeing some stuff drip out over the last year, Astoria (while not cloud, it is a RESTful service that allows us to be cloud ready), Live Mesh (which people confuse as a consumer offering, but actually is a development platform), and SQL Server Data Services (SSDS).

Last week at PDC, Microsoft spoke abut Windows Azure, its cloud services platform. It will consist of web application and service hosting, .NET Services (basically the BizTalk services stack), storage, and data services (SSDS, or now just SDS). Some developers at the PDC were like “this is like the ASP model ten years ago, Azure is doomed to fail.” So the question is, will Azure succeed where the ASP model failed?

The ASP model was about a generic application hosted by an ISP and sold as a service to you. Picture instead of using ADP for your accounting software, you would log onto the ADP web site and use it as a service. This model did not completely fail, but it did not deliver on its mission. It was also a lot of .com hype and about 10-15 years ahead of its time with both the technology and the business acceptance.

While things like Live Services and hosted Exchange is part of Azure, Azure is not about ASP, but about Hosting your app, services, and data in the cloud. There is a need for this: Amazon EC2 and S3 are quite successful, even with the crowd that you think would never put their data in the cloud: Banks. It will take time, but companies will buy into this paradigm and make the shift. The first thing to go into the cloud in masse will be hosted Exchange, then file server capabilities, then applications, then data. Small businesses will go first. It may take years for the shift to be complete, but it will happen. It just makes sense to have your applications hosted in the cloud, why bother and worry about the infrastructure. Infrastructure will be a commodity by 2012. By then most new apps will be hosted in the cloud or using the cloud infrastructure for .NET Services or data.

Only 12 years too late! During the .com era, when I was a CTO of a large .com, I spent 65% of my time worrying about the infrastructure (bandwidth, RAID arrays, load balancing, switches, etc.) Years later at Corzen to support our spidering engines, I focused on infrastructure about 50% (only reason why it was lower than .com era was due to virtualization.) Now I need more bandwidth, more load balancing, it is just a click of a button. Sure it is not going to be this easy, but even if it delivers on 50% of its vision, it will reduce my focus on infrastructure by 95%.

.NET Services (formerly BizTalk Services) in the cloud will get adopted by developers as it matures and as apps get moved to the cloud. SQL Services will get adopted in version 2 when you can do more relational features just as tables, joins, views, etc, instead of the “flexible entities” approach of the first release.

Bottom line is that Azure will succeed, but it will take time for the world to catch up to Microsoft’s vision. Amazon (and to some degree Google) have paved the way.