Edcon, South Africa’s largest nonfood retailer, uses business intelligence (BI) as an important asset in managing its operations. With 50 terabytes of BI growing by 18 terabytes a year, Edcon needed an enterprise-grade BI solution that included
data compression and fast performance. Upgrading to Microsoft SQL Server 2008 Enterprise helped Edcon create an enterprise-grade solution for its BI while gaining increased performance and data compression.
Edcon is the largest nonfood retailer in South Africa, with a share of the South African clothing and footwear market about twice that of its nearest competitor. Since opening its first store in 1929, the company has expanded its footprint to include 1,181
stores under 13 retail brands throughout southern Africa, which helped the company generate fiscal year 2011 revenue of U.S.$3.8 billion. The company also has what may be the largest base of consumer credit customers in southern Africa, with 3.7 million active
private-label credit cards.
||BI is the tool people need to be effective. Our executives rely on our BI system to perform and deliver, which is why we use SQL Server 2008.
| Henri Slabbert
Chief Information Officer
Edcon has long recognized the value of business intelligence, which has played a significant role in helping the company succeed. In 1995, Edcon created its first data warehouse, and in 2001 the company migrated its BI system to Microsoft SQL Server 2000 data
SQL Server 2000 served Edcon well, but the IT infrastructure faced increasing demands with the rapid growth of what had become 50-terabyte data marts. The company also needed to support analytics for its 2.2 terabytes of information hosted in several multidimensional
data cubes. Edcon wanted a database management system that could natively support 64-bit processing, expanded memory, and data compression to boost performance.
On the retail side, Edcon has used SQL Server 2005 for its point of sale (POS) workstations and its in-store databases.
Wanting to take advantage of performance enhancements and new features, in August 2008 Edcon IT staff decided to upgrade the BI infrastructure to Microsoft SQL Server 2008 Enterprise running on the Windows Server 2008 Enterprise operating system. The company
uses Microsoft SQL Server 2008 Analysis Services for managing its analytical cubes, and the SQL Server 2008 page compression feature for reducing the storage footprint of its data. The company is in the process of upgrading more than 1,000 in-store databases
to SQL Server 2008 Standard and 15,000 POS workstations from SQL Server 2005 to SQL Server 2008 Express.
The company uses SQL Server 2008 Analysis Services to manage its 2.2 terabytes of multidimensional data across 20 cubes, the largest of which is 400 gigabytes. About 1,000 merchandisers and analysts access the cubes. Edcon generates some 200 predefined reports
using SQL Server 2008 Reporting Services and distributes them through employee portals created using Microsoft Office SharePoint Portal Server 2003, which it is upgrading to SharePoint Server 2010. It imports information, including third-party data, using
SQL Server 2008 R2 Integration Services, and all Integration Services packages for the POS solution have been migrated to SQL Server 2008 R2 from SQL Server 2005.
Edcon takes advantage of SQL Server replication to move data from POS to headquarters, including information that is imported into the BI data marts and cubes. Within the store, transaction replication sends data from the store server to the POS workstations.
The POS solution uses Microsoft SQL Server Service Broker to queue messages between remote entities. The Service Broker infrastructure processes up to 1.8 million XML transaction logs a day.
Edcon has enjoyed a number of benefits since upgrading to SQL Server 2008, including enterprise-grade support for its mission-critical BI, increased performance, and the ability to compress data.
Gained Enterprise-Grade Support for Mission-Critical BI
Upgrading its BI infrastructure to SQL Server 2008 gives Edcon the enterprise-grade solution its operations require. “BI is the tool people need to be effective,” says Henri Slabbert, Chief Information Officer at Edcon. “Our executives rely on our BI system
to perform and deliver, which is why we use SQL Server 2008.”
Edcon IT has been pleasantly surprised by the demands for BI from across all areas of Edcon. Phil Erasmus, Data Warehouse Manager at Edcon, says, “When we conducted a systems disaster recovery survey with users, to our surprise BI was at the top of the list
along with point of sale, credit, and procurement.”
The company is enjoying increased performance across its BI infrastructure since deploying its new solution, including upgrading to SQL Server 2008 Analysis Services. Ditesh Ranchod, BI Architect at Edcon, reports that complex cube rebuilds to update product
hierarchies after product reclassifications can now be conducted much more quickly. He notes that rebuilds that used to take 36 hours are now completed in less than 5 hours.
“Our product hierarchy updates used to be one week behind the ERP systems because we had to do the product rebuild on the following weekend. Now we can complete the product hierarchy updates by Sunday evening in time for Monday morning reporting,” says Ranchod,
who is part of the Accenture team that helped build and continues to support the BI solution at Edcon. Accenture offers comprehensive BI services to help clients harness structured data to improve decision making, financial management, compliance, and customer
Reduced Storage Needs Through Data Compression
With storage requirements growing by 18 terabytes a year, Edcon needed a data compression solution. By taking advantage of the SQL Server 2008 page compression feature, the Edcon IT staff found that they could reduce the size of their tables by 50–60 percent,
depending on the data sets.
“Compression has meant that our storage facilities are lasting longer,” says Ranchod. “We’re hoping to use compression to reduce our 18 terabytes of annual growth to just 10 terabytes. Additionally, compressed data makes backups go faster and reduces I/O,
which in turn enhances production processing power.”
This case study is for informational purposes only. MICROSOFT MAKES NO WARRANTIES, EXPRESS OR IMPLIED, IN THIS SUMMARY.