Microsoft plans to release its first community technology preview for “Project Madison” in early July.
Project Madison is the code name for Microsoft’s new data warehousing technology for SQL Server. Madison builds off technology Microsoft acquired when it purchased DATAllegro in 2008.
The goal of Madison is to use MPP (massively parallel processing) to deliver high performance and scalability on SQL Server 2008, Windows Server 2008 and industry-standard hardware. The appliance partitions large tables across multiple physical nodes, with each node having dedicated CPU, memory and storage and running its own instance of SQL Server.
The query processing takes place within one physical instance of a database. A Control Node routes queries from applications to all Compute Nodes and then collects and returns the result. Since the data is evenly distributed across multiple nodes and processing occurs in parallel, queries can be performed faster than on single SMP (symmetric multiprocessing) database servers, according to Microsoft.
Microsoft has dubbed the product’s parallel design “Ultra Shared Nothing.”
Microsoft has made a number of announcements around its database and data warehousing business of late. The company recently announced that it was opening up “Project Huron” for early adopters. In addition, Microsoft announced plans May 11 at its TechEd North America conference in Los Angeles to release a CTP of Microsoft SQL Server 2008 R2, aka “Kilimanjaro,” in the second half of the year.
The upcoming version of the database will feature new business intelligence, multiserver management and master data management capabilities, Microsoft officials said.