Microsoft SSDS - Dumb Name, Brilliant Idea

A Google News search for Microsoft SSDS, the SQL Server Data Services it announced last week, yields just six articles – SIX! A random nonsense search term probably returns more than six items.

But SSDS is a big deal, a strategy with far-reaching implications for many parts of the technology industry, deserving of more than six Google News hits. SSDS, as a part of Microsoft’s emerging Web 2.0 superpower strategy serves notice to the status quo vendors, IBM, Sun, HP, Dell, EMC, CA, Symantec and many more, that Microsoft is about to get medieval on them.

SSDS is now in general beta and Microsoft defines it as “highly scalable, on-demand data storage and query processing utility services.” In other words, it’s a giant data storage engine that can be used like electricity or water.

Keep in mind that “data” can mean way more than inventory or contact records in columns and rows of text and numbers. Data can mean documents, e-mails and web pages, or music and video.

Although the pricing details are still a bit sketchy, the SSDS business model means customers will pay for storage as needed, a provisioning model already widely used in network bandwidth and other fiber-level services. Presuming entry-level parity with Amazon, Google and others, the first taste of cloud storage might even be free.

A few weeks ago, Microsoft CEO Steve Ballmer was chatting with the Financial Times and, there, in the mist of the near-future he envisioned a world where “a small number of big companies” would provision computing services from within their own private data centers and manifest those services to desktop, notebook and smart phone users through the Web 2.0 Cloud.

Ballmer made these statements on a press junket in support of the Windows Server 2008 launch, which features virtualization and so fits in well with the cloud computing idea – lots of “leased” copies of Windows Server running on blades in the data center. Ballmer allowed Amazon and Google as examples of prospective cloud providers and this suggests that Microsoft may not be as interested in serving consumers as it is in owning the Business Cloud, at least for now.

SQL Server Data Services, despite its misleading name and somnambulistic introduction, is a big step forward in making the Microsoft Business Cloud a reality

In 1943, Thomas Watson, Sr., the father of the modern IBM, said, “I think there is a world market for maybe five computers.” Change the last word to data centers and the old boy might finally be right.

With only a few big cloud data centers doing most of the computing, it seems likely that the relatively higher-margin market for standalone server gear will be displaced by commodity blades sold like memory chips for cell phones.

The multibillion-dollar systems software market, which includes operating systems, storage and systems management, security, network provisioning and performance management software, thrives in large part on the weaknesses of the pre-cloud computing and networking models. Systems built on client/server, n-tier and Web 1.0 architectures are a dog’s breakfast and have created an industry around software for managing software. Cloud computing still requires system management software, of course, but a different kind. Making money from building it will require a different technical approach and a very different business model.

To understand the potential impact of Microsoft’s cloud strategy and SSDS service initiative on Oracle, IBM/DB2, Sun/MySQL, Ingres, Postgres and other relational or post-relational database management systems, it’s important to note that SSDS has nothing to do with SQL. SQL Server Data Services is a dumb name because, while the service is apparently based at the storage level on Microsoft’s SQL Server technology, it doesn’t use SQL (Structured Query Language) or a variant of it to query the information in the database. It uses a new Microsoft language called LINQ.

SQL was designed to retrieve rows from a tabular structure containing numerical, textual or media elements. LINQ stands for “Language-Integrated Query” and it’s designed to work easily across a wider range of data types and structures than those found in RDBMS products. It does this by making search and query statements a direct part of the programming language in which an application or a service is written, allowing rich programmer-defined data types and simpler coding for data access processes. LINQ enables full text search, page search, and presumably things like picture and audio search against items stored in the Microsoft SSDS database.

Microsoft aims to make money through SSDS by provisioning storage space on a pay-as-you-go basis. Furthermore, it’s making it very easy for standard SaaS applications to be plugged into it through SOAP and REST, enabling ASPs to get rid of their servers and disks by moving their applications up into the Microsoft cloud. The cost of entry is getting rid of the SQL RDBMS, because SSDS sports a whole new “flexible data model” that supersedes the relational model. It’s hard to see what role at all SQL-based RDBMS products can play in the Microsoft Business Cloud. Microsoft may be cannibalizing its own SQL Server business in this strategy but, thanks to MySQL it was already giving it away to all but the largest customers, anyway. RDBMS RIP.

Finally, LINQ is a not a full language in itself. It’s a .NET language extension, intended to bring query capability to the .NET programming languages, C#, Visual Basic and presumably J# and other so-called “CLR” (Common Language Runtime) .NET-compatible languages that may come to be. In other words, Not Java.

The Web 1.0 browser/app server (client/server) model favored by Java and underlying most popular Internet applications servers is on a widely divergent path from Microsoft’s .NET model for network interoperability. Java can certainly remain a viable way to build desktop and browser-based applications that will communicate with the server through Web Service interfaces as they’ve come to do. But, Java’s future as a server programming or service scripting language may be constrained if SSDS is successful in gobbling up major SaaS application providers into its cloud.

Microsoft’s aim seems simple: take servers out of businesses and put businesses onto Microsoft’s servers – think of it as the Exxon of utility computing, the biggest among a few. SSDS might be a tornado, a tipping point, or the flapping of a butterfly’s wings, or maybe it’s just a return to Redmond’s roots – Surround and Embrace, anyone?

It’s virtually a geek cliché to say that Microsoft gets it right on the third try, but that may be what’s happening here. Microsoft Networking Version 1 was a laughable proprietary network flying in the face of the burgeoning World Wide Web. Version 2 was ASP.NET-based web sites and services, a respectable improvement but swimming against the tide of Java and Flash.

Now this. Microsoft’s third try at the network is now, and SSDS, a big part of its Web 2.0 play, makes a compelling proposition: Do it our way and you can get rid of your servers. To most business and IT people, servers equal grief. They’ll listen to this message.

Also, the rest of the tribe is on the mission. Their software development tools in Visual Studio 2008 are up-to-date with the cloud, sporting a lot of new Web Services gizmos and LINQ extensions, and looking further across the Microsoft portfolio, most of it is also poised for a new home in the cloud. Microsoft Office is tooled that way, and so are the Great Plains applications by way of recent Web Services endowments. And, aQuantive was already there when Microsoft bought it, as a family of Web Service monetization products.

It’s been a long time since Microsoft made this much sense.