The final Technology Preview of SQL Server 2008 R2 Parallel Data Warehouse has shipped. This is a significant milestone on the integration of DATAllegro technology as Parallel Data Warehouse now runs on Windows Server and SQL Server.
This Technology Preview offers customers the following benefits:
- High scalability to hundreds of terabytes and high performance through an MPP (Massively Parallel Processing) architecture.
- Hardware choice with appliances planned for HP, Dell, IBM and Bull.
- Comprehensive BI platform with integration to compelling Microsoft BI tools such as PowerPivot, Reporting Services, Analysis Services, and Integration Services.
Early performance tests on Parallel Data Warehouse indicate that the final technology preview is, on average, 45 percent faster than the first technology preview. Microsoft built and tested a Parallel Data Warehouse appliance with up to 50 nodes, 320 cores and 137 TB of physical data storage.
About Parallel Data Warehouse
SQL Server 2008 R2 Parallel Data Warehouse is a highly scalable appliance that delivers performance at low cost through a massively parallel processing (MPP).
A SQL Server 2008 R2 Parallel Data Warehouse MPP appliance acts as an enterprise “hub” that publishes data as needed to various business units or departments (spokes). For dedicated high performance requirements, individual business units can redeploy SQL Server 2008 data marts, or deploy their own Parallel Data Warehouse appliances, as spokes.
In symmetric multi-processing (SMP) architecture, query processing occurs entirely within one physical instance of a database. CPU, memory and storage impose physical limits upon speed and scale. A SQL Server 2008 R2 Parallel Data Warehouse MPP data warehouse appliance partitions large tables across multiple physical nodes, each node having dedicated CPU, memory and storage, and each running its own instance of SQL Server, in a parallel shared nothing architecture. SQL Server 2008 R2 Parallel Data Warehouse delivers improved performance through its unique design as follows:
All components are balanced against each other to reduce performance bottlenecks, and all server and storage components are mirrored for enterprise-class redundancy
A Control Node routes queries from applications to all Compute Nodes, then collects and returns the result
Because data is evenly distributed across multiple nodes and processing occurs in parallel, queries can be many times faster than on single SMP database servers
This architecture can cost less for several reasons:
Rather than relying on expensive proprietary processors or storage, SQL Server 2008 R2 Parallel Data Warehouse appliances use industry-standard hardware
As data volumes grow, scalability simply requires the addition of capacity to the appliance
There is no need for a “forklift upgrade” where the entire appliance must be upgraded
For More Information
- SQL Server 2008 R2 Parallel Data Warehouse
- SQL Server 2008 R2
- Data Warehousing Roadmap Webcast
- SQL Server 2008 R2 Parallel Data Warehouse datasheet
- Data Warehousing 2.0 and SQL Server: Architecture and Vision
- Using SQL Server to Build A Hub-and-Spoke Enterprise Data Warehouse Architecture
Bruce D. Kyle
ISV Architect Evangelist | Microsoft Corporation