WLM is a powerful and flexible tool for managing multiple workloads running concurrently on a single server. With goal-base SLOs, it allows you to tune performance based on the performance characteristics of your applications. If you are running in a vPar configuration, there is even better news. WLM has a global arbiter process known as wlmpard . When running on multiple vPars (within the same server/nPar), wlmpard can automatically migrate unbound CPUs from one vPar to another on the basis that they are being underutilized . The configuration is not terribly difficult. I will point you toward the WLM documentation available at http://docs.hp.com as well as the WLM Web site at http://www.hp.com/go/wlm.