Administrator > DB2

Doing More With Less

DB2 can give SAP users better performance with less cost

Data—it’s the lifeblood of companies around the world. Data runs daily operations, drives business strategy, provides business intelligence (BI) and helps execute ERP. Demands on IT departments are growing annually as companies strive to make the best possible use of the data in the system.

As the database grows, you might think IT needs to boost budgets to acquire more hardware and software to accommodate the rising data load and provide the speed that customers desire. What if you could find a solution that provides better performance, has nearly unlimited scalability, may help delay or reduce capital expenditure, and has a lower total cost of ownership (TCO)? That solution exists—it’s the IBM* DB2* database. System p* users who rely on SAP applications have found that migrating to DB2 has improved performance and lower TCO.

When SAP customers envision the ultimate database, they see a solution that has seamless integration with SAP, superior performance, nearly unlimited scalability, constant innovation, excellent product quality and lower TCO. When SAP customers migrate to DB2, they receive a solution that includes all of those. “We identified three main cost elements: licensing, maintenance and the migration cost including the training of employees to use a new database,” says Ulrich Klenke, CIO, rku.it. “We then compared the offering from our existing vendor with the DB2 Universal Database* (UDB) solution. We concluded that DB2 UDB would cost 40 percent less over five years.”

Compression Reduces Storage Needs

Data requirements call for a system that has storage capacity for the existing data as well as future additions. IT not only must be concerned about online storage costs, but also the cost of multiple data copies for backup and recovery purposes. For DB2, storage is handled more efficiently than competitors with its deep compression technology, which “can cut down the size of database tables up to 70 to 85 percent transparently to the applications, giving the customer better I/O throughput and memory utilization, faster backup and restore times, and reduction of storage needs up to 50 percent,” says Chetan Chaturvedi, Worldwide SAP DB2 Strategy, IBM.

The unique deep compression technology compresses the data on disk. The data remains compressed in memory. “We’re not only seeing significant storage savings. With deep compression you also achieve better memory utilization,” says Chris Eaton, IBM product manager, DB2 for Linux*, UNIX* and Windows*. “If I have 20 GB of real memory, by compressing the data down to only 25 percent of its original size and keeping it compressed in memory, DB2 can effectively store 80 GB of data in the bufferpool for more efficient access and better overall performance.”

This means large savings on buying more memory to support the workload while maintaining good performance. “Our database is now 43-percent smaller than before, and some of the largest tables have been reduced by up to 70 percent,” notes Roland Heim, SAP Basis administrator, INTER Versicherungen. “Despite the compression, there has been no impact on batch performance, and our most important online transactions are actually 20-percent faster with the new version of DB2.” In addition, deep compression is fully supported for all SAP releases and applications, from R/3 3.1l up to the newest releases.

Unique DPF Scale-Out Technology

Another cost-saving DB2 feature is its Database Partitioning Feature (DPF), which offers a scale-up and scale-out technology for SAP BI. Fully supported since SAP BI 2.0, most larger SAP BI installations on DB2 use DPF.

“DPF allows the customer to do more with less,” notes Chaturvedi. “From an SAP perspective, DPF is an advantage. It offers the client the best exploitation of the available hardware through logical and physical partitioning. It allows customers to run high-end SAP BI on a lower-cost hardware and to deploy it on multiple smaller boxes instead of one big SAP machine,” he says.

Advantages of Multi-Dimensional Clustering

Multi-dimensional clustering, or MDC, is yet another unique DB2 feature allowing customers to do more with less. Instead of using range partitioning, MDC uses clusters along multiple partitions. “The customer can apply this to as many dimensions as desired,” Chaturvedi says, pointing out that range partitioning primarily works in just one dimension.

“We use DB2 as the database of choice for our SAP NetWeaver BI solution because it offers simple administration and lower licensing costs,” says Matthias Assmann, head of management, Information T-Mobile UK. “System performance in parallel mode has increased, and the fully integrated database cockpit helps to reduce operator workload.”

Next page: >>