DAT 101: Microsoft Data Platform Vision
It is my first day of TechEd 2006, and I’m starting with a presentation by the data team. The deck is entitled “Microsoft’s Data Platform Vision” and the initial speaker is Dave Campbell. I see Gert Drapers, Alex Payne and Pablo Castro in the room too. Sharon Dooley is here too. Dave framed this as a “where we are going talk.”
· Standard Kilo to Mega, Mega to Terabyte introduction
· Amount of data being “born digital” continues to grow
· The concept of permanently persistent data for prediction in sales and marketing becoming strategic assets
· Concept of pre-transactional data (record of intent) with less active management
· Made an interesting comment about spatial data and this becoming more important
· No longer about the database, it is about the higher level services.
· Latency and human costs are the drivers now
· Time to solution is very important
· Next frontier is automation of process interfacing with people
· Talking about bleeding device interfaces and dynamic applications
· Katmai goal is scale out and always on-line with better health monitoring – “Grid for real.” Policies for operations, policy for data. Semantic data models controlling aspects like security of data.
· Announced Always-On program. Again.
· Image acquisition and structured information with workflows and data mining
· Clearly giving the sense that spatial enablement is critical
· Greater emphasis on conceptual data modeling, target for entity framework. Unification of data models for use and programming. Clearly schema becomes very important
· Talking the insanity of dynamic SQL from the side. The idea is to have LINQ enabled data sources instead. Don’t program against the API, solve the business problem. LINQ over dataset (already announced) and DLINQ for SQL (already announced) and LINQ over entities. Also talking about a new XML data storage format for VS.NET
· Dave started picking playfully on Pablo for his typing skills
· Pablo gave a pretty good demonstration of the basics of DLINQ.
· Alex Payne gave an interesting demo of Excel consuming Analysis Services cube and consuming data by schema. Drill-through mostly using the new pivot table stuff.
· Talked about Federations of data sources and disconnected data sources and data being model driven and SOA.
· Talked about SQL Mobile as the foundation of SQL Server Everywhere Edition and the CTP is available today. Session DEV343 for drill-down on that.
· Cameron talked about VSTE for Database Professional (aka Data Dude)
o Schema management toolset
o Allows to delta and test in offline mode
o Showed the dependency walker for meta-data, very cool
o Showed the schema differencing tool. Very nice.