On the client side a Job is created and one or more Node implementations with the
required data (configuration / data streams) are assigned to it. This job then
contacts the server and queues the nodes there. Thanks to the communication
interface being asynchronous, the client is not blocked while the server
processes the data.
The nodes form a directed, noncyclical graph, which defines the workflow (e.g node 1 loads data, node 2 processes it, and node 3 sends it back to the client. the section called “Application Scenarios and Code Examples” elucidates how to implement this and offers typical examples for illustration.
