Utility Computing Through Virtualization Management

A leading Web portal company was faced with the daunting challenge of constantly needing to reallocate computing capacity quickly to handle dynamic changes in the loads of the various content engines of the portal. For example, important world events typically placed high loads on the news and finance content engines, while bad weather resulted in high loads on the weather and travel content engines.

It was impractical to provision enough computing capacity to handle the aggregate of the peak loads of each content engine. To do that would have doubled the company’s capital costs and tripled the energy costs associated with the data center. Instead, this company turned to virtualization to allow more efficient use of its existing servers and faster provisioning time. Much of this process involved manually interacting with VMWare’s Virtual Center. This process was typically initiated when users complained of slow response time and usually took three to four hours to provision the appropriate virtual computing required to reduce the stress on the heavily loaded content engine.

The company selected Optinuity to implement an Autonomic Policy that would automate the dynamic provisioning/allocation of computing capacity – before users were impacted. With Optinuity Oasis™, this company implemented an Autonomic Policy that continuously monitored key performance metrics of each content engine and analyzed those metrics for significant trending against predefined baselines. When certain trending thresholds were crossed, the Autonomic Policy would interact with VMWare to automatically provision or de-provision virtual capacity and then make that computing capacity available to the appropriate content engines.

Results:

• Achieved utility computing through the combination of virtualization and Autonomic Policy Management
• Saved millions of dollars per year in capital and energy costs