For a long time, capacity planning has been directly linked to the scarcity and cost of resources. In the mainframe era, it was considered a necessity. In the distributed system world, it faded when hardware resources became seemingly abundant and cheap. The market is now evolving from a scarcity-cost equation toward a quality-cost equation: Capacity planning is a solution to maintaining business service quality and avoiding the consequences of downtime and brownouts — and no longer a way to maintain a minimum level of service at the lowest possible cost. As data centers struggle with server consolidation and server virtualization, capacity planning becomes the key to maintaining or improving service quality while containing costs.