PeopleSoft Integration Broker provides a mechanism for communicating with the outside world using XML files.

Communication can take place between different PeopleSoft applications or between PeopleSoft and third-party systems. To subscribe to data, third-party applications can accept and process XML messages posted by PeopleSoft using the available PeopleSoft connectors or by adding a custom built connector to the Integration Gateway.

This topic primarily covers publishing outbound asynchronous messages from a PeopleSoft system to a third-party application using the delivered File Output connector.

For information on outbound synchronous messages, see “Using Outbound Synchronous Messages”.

To send a message, you must properly configure various internal structures and processes.

Here are some of the relevant components:

• Gateways are essetially a pathway out of the system. There is always a LOCAL gateway in PeopleSoft which refers to the source PeopleSoft system. Other gateways may exist – for example when there are differences between PeopleTools versions.
• Domains refer to PeopleSoft application server (tuxedo) domains.
• Nodes define what application a message belongs to – any system that integration broker can talk to (including the local system) will have a node.
• Messages define the structure of the XML sent between nodes. The key difference between asynchronous and synchronous messages is acknowledgement – synchronous messages need to be acknowledge by the receiver, aynychronous ones don’t. Synchronous messages aren’t used much in PeopleSoft, but I believe their main purpose is for reliability.
• Queues are used to manage and group messages so that all messages are sent through the multiple pipes. Queues are used for administration of integration broker to avoid stopping everything when something is wrong. For example you can pause a one queue for messages to the report architecture without say stopping messages to the finance system.
• Service Operations bring together messages, queues, handlers and routings. They essentially define how the message moves between systems.
• Services seem to be used to group service operations.
• Routings are used to link a service operation to a node, and can also specify transformation of a message.
• Handlers contain code and logic for sending/receiving/manipulating messages.