IBM Offers Linux Clustering Support for DB2

New offering, aimed at medium-sized enterprises looking to grow, lets customers scale systems from two to 1,000 nodes.

IBM Monday announced and demonstrated at LinuxWorld a new Linux database clustering solution for DB2 users that will allow businesses swamped by e-business-generated data to enter into the world of multi-terabyte without being shackled to proprietary systems.

IBMs DB2 for Linux Clustering solution is an end-to-end offering that enables customers to scale systems from two to 1,000 nodes, adding nodes as required.

According to Jeff Jones, IBMs director of strategy, Data Management Solutions, the offering is based on a building-block concept that provides a growth path that can potentially encompass both IBM and non-IBM software, including IBMs WebSphere e-business application server, its Tivoli systems management software, as well as SAP R/3, mySAP, and mySAP Business Intelligence.

Jones, in San Jose, Calif., said that the building-block approach is the first of its kind. Until recently, he said, customers had to predetermine how much scalability had to be built into their systems as well as having to then build them on top of proprietary systems. "This brings a whole new … simple way to get started," said Jones.

The move is aimed at targeting medium-sized enterprises that are looking to grow, Jones said, including financial, retail, manufacturing and public-sector industries. Jones said its also ideal for enterprises requiring highly scalable, low-cost Linux clusters to power ERP, CRM or business intelligence projects. "Whats behind this is the fact that the rate and amount of data being collected is not slowing down," he said. "The drivers are only accelerating, as businesses continue to get real good at collecting all sorts of data. Of course, the cost to collect this data is more urgent than ever."

On the high-end scale of clustering lies customers in the data warehousing space, including engineering and scientific enterprises with enormous banks of information, such as genome projects and protein-mapping projects, or high-volume transaction processing clients, such as American Express. On the low end, IBM is looking at signing up customers that have outgrown a large, single server and are looking to take the next step.

According to Jones, scaling up is now easier than ever, thanks to simplifications in the hardware end of things. "Adding a node isnt something mom-and-pop grocery stores would do overnight," he said. "But on the hardware end of things, doing it has become greatly simplified. Its only a few cables, instead of dozens of cables to connect the new computer into the other servers in the cluster."

A small, two-node configuration sells for under $10,000 U.S. It includes two IBM xSeries eServers (model x335), DB2 Workgroup Server Edition for Linux, SuSE Linux 8 and additional application components. IBM business partners including SAP, SuSE, Voltaire, Mellanox, InfiniSwitch and TopSpin are offering the blueprint for the solution and a cluster configuration overview to assist customers with configurations.