When work enters the ODR, an optimization effort is performed to achieve a desired balance of performance results. As the work flows, the DWLM (Dynamic Workload Manager) performs load balancing on it. As work variations change, and the balance of work in the nodes becomes distressed, the application placement controller then rebalances running applications in order to ensure efficient workflows.

Optionally, WebSphere Extended Deployment provides the provision to hook-up with TIO (Tivoli Intelligent Orchestrator). TIO is used to add and remove machines from the pool of resources being managed by WebSphere Extended Deployment.

For example, if WebSphere Extended Deployment determines that it can't meet transactional volume, Tivoli Intelligent Orchestrator can provide more server hardware on which to manage WebSphere applications. Tivoli Intelligent Orchestrator can then allocate more server resources to WebSphere Extended Deployment, which will add the resources to the available pool and balance the workload across the new, larger set.

Extended manageability and high computing
Extended manageability helps to improve the management of complex WebSphere deployment operations with real-time, advanced visualization tools and gradual, controlled implementation of autonomic capabilities. This visualization capability enables the administrator to understand the operational state of the environment.

The extended manageability GUI will not only allow administrators to view the current state of the environment; it will allow some administrative action to be taken. For instance, an administrator will be able to override decisions made by the provisioner, or modify application priority through the operations console.

High performance computing support comes with the high availability manager, and the WPF (WebSphere Partitioning Facility). While these are separate components, the highly available infrastructure is very much a part of the application partitioning ability of WPF.

The WebSphere Partitioning Facility supports application partitioning, whereby applications divide logic and data into subsets (e.g., by ranges of customer numbers) that can be mapped to partitions in servers and in databases, thus increasing transaction throughput and reliability.

However, in order to gain the benefits of partitioning, the entire Extended Deployment environment must be aware of the partitions and WebSphere application has to build with WPF components.

Also, an additional component called HA Manager is available to monitor the server environment, and in the event of server failure, provide application server recovery or routing to another available server. HA Manager supervises the execution of JVM's (Java Virtual Machines) in the cell.

The cell containing the JVMs is a member of a Core Group. A Core Group is a new term introduced by the HA Manager component. A JVM within a cell is a member of exactly one Core Group. Additionally, a cell consists of at least one non-overlapping Core Group of JVM's.