Whether you are adding a new system to your landscape or integrating your existing systems, there are three key things to figure out before starting your integration project: (1) the types of systems you want to integrate, (2) which protocols are used to communicate and (3) what data needs to be integrated. Read below why our integration experts recommend you find an answer about these topics upfront.
Why is knowing the type of systems you want to connect with each other so important?
A lot of systems are written in a different programming language. Each of them has their own communication layer that is kept as close as possible to the core of the programming language, due to the various implementations and protocols being used to translate data in very different ways. Therefore, it is important to know if you are connecting cloud to cloud, on premise to on premise or a combination of cloud and on premise.
If you have the answer to this first question, you can dive a bit deeper and start investigating which protocols are used for communicating between systems. Do your systems use an API (REST, SOAP, etc.) to expose the functionalities or are they using specific protocols like MQ, I-Docs, etc.? The different protocols that are used between the technologies have their own implementation to translate data in a certain format, but also the way of pushing/receiving between systems is different. Implementing a REST API is completely different compared to an IDOC. Not knowing this upfront might cause a lot of useless integration scenarios or even require a restart from a blank integration. Usually, you can find these protocols back in guidelines and best practices shared by the company that distributes the system. If they are not shared, you might need to activate the communication technology and check the kind of structure provided by the system itself after activation (WSDL, EDMX, etc.).
Finally, you need to know the format and structure of your data. It is not because you manage to use the suited protocol that the systems can interpret each other’s messages correctly. Translating the structures, mapping the values, and serializing the messages are some of the main challenges in every integration project.
The structure of the data is written in the metadata provided by the system where communication is coming from and going to. SOAP for example has a WSDL, OData translations work with EDMX files, etc. When working with REST Endpoint, a swagger can be shared containing examples of the message itself, but also possible answers and errors.
Currently, when structures are changing on node level, the mapping has a lot of impact because the position of certain fields is lost. That is why it is very important to work closely together with different teams during an integration project. Communication between systems might be important, but the communication between integration teams might be even more important to get a working solution!
If you can answer these three questions upfront for the systems that you want to integrate and ensure good and transparent communication across teams you are on your way to success.
Having trouble answering any of these questions? Our integration experts would be happy to help you on your way.