While there are several exciting new features included in SQL Server 2008 R2 including PowerPivot for Sharepoint and Excel, Master Data Services for data consistency and StreamInsight for processing high volumes of event stream data, what I am most excited about isn’t a core new feature but one of two new editions that are being introduced – SQL Server 2008 R2 Parallel Data Warehouse (PDW).
Parallel Data Warehouse is an appliance based solution using vendors including EMC to provide a highly scalable data warehouse. PDW uses a massively parallel processing (MPP) architecture that scales from terabytes to petabytes while delivering high performance and using SQL Server additional business intelligence features and functionality.
So what is a data warehouse and why does an appliance based solution make sense? For applications that are transactional in nature (OLTP), you can easily dedicate a set of disks to satisfy the small random access of data. If you run into performance problems you can devote more resources to the server to quickly address any issues. With Data Warehouse environments, there are large amounts of data that are being stored over a long period of time with analytic processing requirements. Having the capacity to store this data as well as having adequate performance for these requests is what makes this solution compelling.
But an additional gain to the appliance solution is the ability to provide parallel management of these large volumes of data. With parallel processing, data is stored on more than one device so that all of the resources in the appliance can access and process the data at the same time. If a server doesn’t have enough resources to process a request then multiple servers can be used together to process the load.
By including EMC’s industry recognized high performance and highly available arrays into the SQL Server 2008 R2 Parallel Data Warehouse solution you have a dedicated appliance that has the power of Microsoft and EMC to handle some of the largest most complex data warehouse environments.