Technical Support Services Blog
Discover the latest trends in technology, and the technical issues customers are overcoming with the aid of HP Technology Services.

HP AppSystem for Microsoft SQL Server 2012 Parallel Data Warehouse: What's all the noise?

by Kevin Lange, Master Database Architect

 

When it comes to Parallel Data Warehouse (PDW), the 1st order of business has always been blazing compute and IO performance.  PDW has been synonymous with speed since the initial integration with Data Allegro, v2.0 pushes that envelope with updatable column store indices.  A column store index stores data in [horizontal] columnar format, allowing for much more of the index read data to the stored on the same 8k page; SQL data is read in these 8k chunks, less reads more data.  With Columnar indices,  only the data actually required to satisfy the query is transferred from disk to memory [not the data of other columns like with traditional indices, disk pages containing the other columns data are simply ignored]. In doing so, the number of pages transferred from disk to SQL memory is tremendously less.  v2.0 applies aggressive compression algorithms to further reduce the disk I/O needed to serve the query request and yield up to 50x gain in query performance.

 

Capacity is now increased to multiple petabytes per BI implementation, again due to increasing compression, as well as improved process and thread parallelization and process dispatch efficiencies.  6PB + capacity places a Parallel  Data Warehouse in the sweet spot of enterprise class BI data operations, right out of the box, with the familiar HP SQL Server services  [which include; Database Consolidation Reference Architecture, Enterprise Data Warehouse, Fast Track Services, Oracle to SQL Server Migrations, Performance Tuning and High Availability Clustering and BI Workshops and Implementation Services] and tool set [SQL Server Analysis Services, Integration Services, Reporting Services, Master Data Services and Data Quality Services].

 

Extensibility stems from the integration PDW2.0 now has with non-structured data stores [like Hadoop].  A data interoperation technology coined PolyBase, provides a unification layer between [Hadoop Cluster] MapReduce calls and traditional SQL calls.  Simply put, a BI developer can now write a connected SQL query that can access a non-relational Hadoop data structure and retrieve that data, mapped to a tabular format.  Even more exciting is the fact that PolyBase allows for joining existing SQL Server based relational tables to the mapped non-relational data, in effect integrating both realms of customer information.

 

Finally, Total Cost of Ownership; TCO has always been a mainstay of value for SQL Server deployments.  In a study published with the V2.0 release; EMC, IBM, Microsoft, Teradata, and Oracle Data Warehouse deployments were independently compared for capacity and speed, both pivoted against equivalent price.  The bottom line, HP PDW V2.0, substantially outperformed the competition.  SQL Server's $5 per TB of data was almost half the cost of the nearest competitor.  http://www.valueprism.com/resources/resources/Resources/PDW%20Compete%20Pricing%20FINAL.pdf

 

PDW V2.0 is indeed a game changer which is generating significant industry buzz, as the final production bits are being shipped.

 

As someone who has been around data management for some time, I can say that PDW V2.0 is a great product.  As good as the product is, customers should know that their support experience with PDW will be a great because HP delivers the services that drive performance.   From our Assessment, to ensure that  the  system is planned and well-sized,  to our Deployment Accelerator and Advisory services, to ensure that we have a  smooth transition  into production  -   our HP/Microsoft unified approach  to Appsystem support for PDW, ensures that issues are resolved quickly and without any confusion.

 

For more information, please visit our website @  HP AppSystem for Parallel Data Warehouse.

 

 

KevinJune.jpg

Kevin Lange is a Master Database Architect, and has over 27 years in enterprise information systems.  Kevin has a developer's passion, with skills spanning traditional client-server architecture, virtualized and cloud/web-enabled deployments, application development and  IT infrastructure, and database design, architecture and deployment.

Leave a Comment

We encourage you to share your comments on this post. Comments are moderated and will be reviewed
and posted as promptly as possible during regular business hours

To ensure your comment is published, be sure to follow the community guidelines.

Be sure to enter a unique name. You can't reuse a name that's already in use.
Be sure to enter a unique email address. You can't reuse an email address that's already in use.
Type the characters you see in the picture above.Type the words you hear.
Search
Showing results for 
Search instead for 
Do you mean 
About the Author


Follow Us
The opinions expressed above are the personal opinions of the authors, not of HP. By using this site, you accept the Terms of Use and Rules of Participation