Are Blue Skies Ahead for Azure?
By Jonathan Goodyear
If you've been monitoring Microsoft over the past few months, you know one of their big new ventures out in Community Technology Preview is the Azure Services Platform (www.microsoft.com/azure/). Azure Services Platform (Azure for short) is a cloud computing environment that provides a scalable hosting solution for all or part of your application. Microsoft's vision is that some companies will host entire applications with Azure, while others simply will utilize some of Azure's services, such as data storage and synchronization, workflow, service bus, federated identity, access control, SharePoint, Dynamics CRM, etc.
The basic idea behind Azure is that many companies need large scalability, but only some of the time. For example, if a company is featured in a news story that gets them on the homepage of Digg or SlashDot, their website and/or services will experience a temporary high influx of traffic and usage. Azure would enable on-demand scalability to handle the additional workload. A second example would be a seasonal retailer that does most of their business around the holidays (think flower vendors or toy companies). Azure could help them scale their ecommerce operations to handle holiday sales, without them having to make a large capital investment in infrastructure to cover relatively short bursts of application activity. Start-ups also could leverage Azure to get up and running quickly and (in theory) inexpensively, thus saving their limited financial resources for other expenses, like product development and marketing. Azure does a lot of other things, but I'll defer to the Microsoft website to fill in the gaps.
While Azure has an impressive feature set, it also faces some considerable challenges, which I'd like to discuss here.
Distrust of Microsoft. Microsoft has done a lot to improve its image since it announced in 2001 the Hailstorm service, which offered many of the same features. However, there still is an undercurrent of distrust among many in the industry. It will be critical for Microsoft to avoid any gaffes that could undermine the progress it has made, because Azure's success is riding on how trustworthy Microsoft appears when it launches.
Uncertainty of Launch Date. A lot of companies will not even consider Azure as an option until Microsoft releases a better indicator as to when the service will be released. Microsoft has long held the philosophy of "we'll ship it when it's ready", which is a good strategy to maximize quality. However, given the long planning phases of application development in larger companies, this could push out the time frame for when Azure sees a ramp up in usage.
Existing Competitors. Speaking of an uncertain launch date, there are some formidable competitors in the cloud computing services platform space. These include Amazon's Elastic Compute Cloud (Amazon EC2) and The IBM Cloud (aws.amazon.com/ec2/; www.ibm.com/ibm/cloud/). Granted, these two offerings are much more limited in their scope of functionality, but both are here today and have very trusted vendor names behind them.
Reliability. Microsoft has done a lot of research and development to create what they claim is an extremely reliable and fault-tolerant fabric into which your applications can be woven. The CTP has already had a few hiccups in this realm, but that's why it's a CTP and not RTM. This issue may end up being a red herring, but Microsoft doesn't always have the best track record of reliability for version 1 offerings. They'll need to nail it with Azure to survive and not fall in to their typical roadmap of three versions to get it right.
Fear of Outsourcing and Data Loss. Larger companies have become increasingly less afraid of outsourcing certain operations, such as CRM. There are still lingering fears in many industries, though, such as medical, financial, and government. It will be interesting to see how Microsoft handles concerns over HIPAA compliance, intellectual property protection, and international borders (e.g., where data is not allowed to be stored outside a particular country's geographic boundaries).
Unknown Pricing Model. It stands to reason that Microsoft is going to do a lot of research to come up with a pricing model that is affordable to the customers they are trying to attract. But the exact pricing model is still unknown. Until some light is shed on pricing, it will be impossible for companies to begin the budgeting process. Microsoft also has some very attractive options for start-ups, including its BizSpark program (www.microsoft.com/bizspark/), which appears to conflict in some ways with the objectives of Azure. After all, free software licensing is hard to argue with, even if you're footing the bill for the hardware. If Microsoft does a good job of creating attractive pricing options for companies both large and small, though, Azure could be a big hit.
It's Not "Exactly" Plug and Play. Azure also will allow developers to leverage their existing skill sets with Visual Studio and (in the near future) other tools and languages, such as Eclipse, Ruby, PHP, and Python. However, it's not plug and play. You can't simply upload an existing application to the Azure cloud and take advantage of infinite scalability. Rather, Microsoft has exposed a set of interfaces into its services that are both RESTful and SOAP-compliant, making them easy to use. In addition, Azure has a couple of different data storage offerings, but you don't access your data using ADO.NET as you're accustomed. You'll need to architect your application to be Azure-compliant to get the most benefit. Companies looking for a quick fix for their existing poor-scaling application are not going to find it with Azure. Some work will be involved, but the upside potential of putting in that work is enormous.
Like any new technology, Azure promises to solve an existing challenge; in this case, application scalability, with the added benefit of some key additional service offerings. It s a large undertaking that very few companies could even think of attempting. Fortunately, Microsoft is one of those companies that has all the ingredients to make Azure successful. How they fare in their actual execution of that vision remains to be seen. I'm really pulling for them, though, because whether or not they succeed, the needs they are aiming to solve will remain.
Jonathan Goodyear ([email protected]) is president of ASPSOFT (http://www.aspsoft.com), an Internet consulting firm based in Orlando, FL. Jonathan is Microsoft Regional Director for Florida, an ASP.NET MVP, a Microsoft Certified Solution Developer (MCSD), and a contributing editor for asp.netPRO.