COCOP Communication Stack
The COCOP communication stack was developed in the research project COCOP (Coordinating Optimisation of Complex Industrial Processes). Its intention is to enable a decoupled and scalable architecture in industrial software systems. This page provides a brief introduction.
Structure
The COCOP Communication Stack consists of two layers, namely standard-based message structures and AMQP as the communication protocol (see the following figure). These enable software applications to exchange information regardless of the runtime or the implementation platform. Such interoperability is an important factor in an industrial environment, as systems often come from multiple vendors. In such an environment, the COCOP Communication Stack can interconnect any vendor-specific interfaces with an appropriate adapter. Although adapters generate overhead in design, there is no need to support multiple interfaces in any system, because the single adapter enables compatibility with any system connected to the message bus.
In the stack, a core principle is that the layers do not have any dependencies to each other, which adds flexibility to design. It is possible to modify either of the layers without affecting to other. Therefore, if desirable, you can introduce new message types for communication or add another communication protocol, such as HTTP or MQTT.
Communication Patterns
Publish-subscribe
COCOP favours the publish-subscribe communication pattern implemented with a message bus. In this pattern, data is streamed from the data producers (i.e., data sources) and routed to data consumers based on a topic. That is, data is ‘pushed’ from the source rather than ‘pulled’ by the consumers. This pattern does not fit well for all communication needs but excels in scenarios where data is frequently updated. In such scenarios, publish-subscribe reduces network traffic, as the consumers do not have to poll for updates. On the other hand, the pattern facilitates the design of event-based systems where timely reactions are expected from network nodes.
Request-response
Because publish-subscribe cannot meet all communication needs, the request-response pattern is also supported in COCOP. This need stems from the fact that sometimes the user of data must initiate communication. For example, consider a situation when you need one or more measurement values for a particular moment or time period in the past. On the other hand, you may want to request a network node to perform some action (such as a long-running simulation task). In this pattern, data is ‘pulled’ by the consumers rather than ‘pushed’ by the producers.
Because a message buses (RabbitMQ, in this case) do not consider request-response the primary communication pattern, COCOP has implemented a dedicated API. Although this functionality does not require lot of coding and the RabbitMQ website also provide an example, the API saves some work and enables the logic to be re-used.
Acknowledgement
COCOP - Coordinating Optimisation of Complex Industrial Processes
https://cocop-spire.eu/
This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 723661. This piece of software reflects only the authors’ views, and the Commission is not responsible for any use that may be made of the information contained therein.