The partitioning facility (WPF) provides the capability to partition HTTP requests across back end WebSphere Application Server instances. Known as HTTP partitioning, this capability works with the on demand Router (ODR), which uses an awareness of partition location to forward HTTP requests to the appropriate target application server. This section provides an overview of the HTTP partitioning function, the architecture of the solution, and defines key concepts that are required to implement HTTP WPF applications.
The HTTP WPF partitions are no different than Enterprise JavaBeans (EJB) workload Internet Inter-Orb Protocol (IIOP) based partitions: they correspond to a related collection of data elements that are commonly accessed together. They have the distinct feature that no data element belongs to more than one partition. This property enables Web applications to make inferences in how these data elements can be treated. That is, no data element belongs to two partitions and a single partition is only active on one application instance at most.
When an application receives a request for a particular partition, it can be certain that no other application instance is accessing this partition and the associated data elements. As such, the application can leverage this certainty by being more aggressive in treating the partitioned data. When combined with techniques such as caching and batching, this application performance is improved by increasing the efficacy of size-limited caches and reducing the overhead in ensuring data consistency.