Arbortext PE sub-process Pool Management
A subprocess pool is a collection of identical Arbortext PE sub-processes. Each Arbortext PE sub-process pool has a unique name, a collection of attributes that describe the pool's behavior, a set of parameters that are passed to every Arbortext PE sub-process in the pool, and a collection of rules that determine which HTTP requests can be processed by Arbortext PE sub-processes in the pool.
Support for multiple Arbortext PE sub-process pools offer two major advantages:
• Pools of Arbortext PE sub-processes can have different parameters, making it possible to have groups of Arbortext PE sub-processes with different behaviors.
• Arbortext PE sub-processes pools can be configured to reserve performance capacity by ensuring that requests of a particular type don't consume all available Arbortext PE sub-processes.
Each Arbortext PE sub-process pool provides the following capabilities:
• ensuring that a configured minimum number of Arbortext PE sub-processes are running.
• starting additional Arbortext PE sub-processes (up to a configured maximum number) upon request.
• limiting the number of Arbortext PE sub-processes that can be used to process queued transactions.
• terminating any Arbortext PE sub-process that fails to respond to a request from the Arbortext PE Request Manager after a configured period of time.
• terminating Arbortext PE sub-processes that have been idle for more than a configured maximum period of time.
• terminating an Arbortext PE sub-process and starting a new one (between requests to avoid disruption of processing) if an Arbortext PE sub-process has been running for more than a configured maximum period of time.
Terminating an Arbortext PE sub-process avoids possible consumption of resources that could accumulate over time.
• returning messages that a request has failed after a configured interval if all Arbortext PE sub-processes are busy and the maximum number of Arbortext PE sub-processes are running.