Big Blue has developed "self-optimizing" software that will let businesses more flexibly react to spikes in computing demand. The software would, for example, let a corporate Web site automatically bring more computers online to respond to an unusually high number of visitors.
Typically, companies buy an excess of servers, software and storage to handle peaks in processing demand, which, however, may only occur on a monthly basis--when processing regular financial reports, for example. The self-optimizing software from IBM is designed to let companies buy only what they need, executives at Big Blue said this week.
With IBM's forthcoming software, "you can provision a normal amount and, with the changes in traffic that are forecasted, you can dynamically bring resources online and down again," said Ric Telford, director of architecture and technology, autonomic computing at IBM.
Big Blue will be showing off its latestsoftware next week at the CeBit trade fair in Hannover, Germany. The demonstration, which IBM is calling dynamic system adaptation, shows how an airline Web site can allocate and then bring down WebSphere server software to meet predefined response times.
The self-optimization software complements other autonomic computing technology IBM has developed, including its self-configuring,and self-protecting software.
IBM will first use the self-optimization software as part of its consulting services and later embed it into products. Likely recipients are IBM's Tivoli systems management software, as well asand server products, Telford said.
IBM competitors Sun Microsystems, with itsplan, and Hewlett-Packard, with its Utility Data Center, have launched their own projects intended to increase use of customers' existing computers.
There are three components to IBM's autonomic self-optimization software, including what IBM calls adaptive forecasting, rapid reconfiguration and online capacity planning.
The software can study traffic patterns to predict changes in computing demand and then automatically bring more resources online to handle the changes in computing workload. The software can also generate historical data to analyze trends over time.
Based on customer feedback from its consulting engagements, IBM will enhance the self-optimizing software, the company said.