Below given is the architecture of Publish/Subscribe coordination based model. Here, you can see multiple processes are connected by a broadcast network. For the sake of simplicity processes are marked as Process 1 to Process n. Every process in a Publish/Subscribe coordination based model can act as a producer of information, consumer of information and both a producer and consumer of information. A producer process is the one which creates information and a consumer process is the one which uses the information.
Publish: Whenever a producer process has new data, it is broadcasted over the local area network (LAN) in the form of tuples. Then, the broadcasted tuples are copied in the RAM of tuple daemon. After that, a copy of this information is sent to whichever process has subscribed to this information. This method is known as publishing.
Subscribe: A process which requires this data subscribes to this information. This method is known as subscribing. Whenever a process subscribes for information, it has to be informed about the new updates.