Financial giant Nedbank has managed to dramatically reduce the time to information of its Business Intelligence (BI) and Analytics systems, with the deployment of SAS Scalable Performance Data Server (SPDS).

Nedbank was facing a challenge in that it was taking in excess of 13 hours to access and draw data from SAS datasets in order to generate a mart from a star schema, which the users would then use to do queries with, via Enterprise Guide. This translated to a massive delay in reporting, making real-time BI and Analytics just about impossible. After calling on its BI partner SAS Institute to investigate the issues that may be causing the problems, SAS suggested that they try deploying the SPDS technology in the Nedbank environment.

"In all honesty we were asking a lot from a relatively underpowered technology environment, but when it takes more than 13 hours to get to the data you need to, its just not intelligence," says Christo Toerien, Group Technology Executive: Data Warehouse Infrastructure at Nedbank. "That said we called on the team at SAS to come and help us formulate a solution, using the technology we had, without reinventing and redesigning the entire system."

The Challenge
With time to data from the Nedbank Debt Manager being what it was, SAS and Nedbank had to bring the time to construct the data view using the ETL tools down. Some challenges facing the project included the fact that both the Analytics and BI tasks had to be done off the same server.

Says Nicholas Eayrs, Manager - EMEA Solution & Technology Innovation Centre at SAS Institute Global: "The Nedbank situation was not unique but it was challenging. The company currently runs IBM pSeries hardware and the particular system we were working with was running on the back of a SAS dataset, and needed the capability to be able to update a dataset while it was being queried. We had a quad processor dual core partition to work with and needed to reduce the time to intelligence while at the same time ensuring that we didn't impact the Total Cost of Ownership of the infrastructure footprint."

The technology
With very little time at hand, it was identified that Nedbank already had SPDS as part of the Old Mutual Group enterprise licensing agreement, and this could be used to address the data access issues. The team quickly swapped the data store from the SAS dataset to the SPDS subsystem. Instead of creating a separate mart, a view on the star schema was created reducing the previous 13 hours it took to transact with the system to mere seconds. All of this while at the same time keeping the high performance we required from a reporting perspective.

SPDS is a native tool to SAS, which serves large numbers of concurrent users through the use of parallel processing and data server capabilities, it offers both vertical (user and queries) and horizontal (data volume) scalability. The technology is optimised to deliver subsets of information that need to be harvested from large enterprise data mountains on demand.

The solution
Says Adrian Mattioli, EDW Infrastructure Manager at Nedbank: "SPDS was rolled out in as little as three days for development and as little as 10 minutes for Query and Analysis and Production once the configuration could be duplicated this dramatically reduced the amount of time it takes for us to access our data, and what is even better is that we have not had to purchase any additional hardware or software to make it happen. It truly completes our Intelligence Storage Offering."

When installed SPDS simply embeds itself and just acts as an interface within your system, integration is seamless and it can work with just about any open standard database. Nedbank is now able to take advantage of the fact that its BI and Analytics applications maintain consistent performance and that ETL processes do not exceed the time available as the amount of your enterprise data continues to grow.

Benefits include:

• Significantly speed up the gathering of subset information with the use of parallel storage technologies and hardware.
• Optimised performance for both business and analytics applications, through a single integrated platform.
• Reduction in extraction, transformation and loading times.
• The ability to connect to data from different servers
• Data compression has enabled it to save storage space and also speed up processing.

To end
"Nedbank was given a fully functional system which supports the business aims of the organisation, in just on three days. It is cases like these which highlight the fact that sometimes the intelligent use of what you already have, can solve your technology challenges without having to spend more money on additional software and hardware to support your business aims," ends Eayrs. "Its all about more for less and bridging that gap from time to intelligence."