Traditional streaming solutions suffer from one of two issues. In the case of pulling, there is a need for locking or extensive use of resources on the side of the quick consumer. In the case of pushing, there is a possibility that a number of messages to process will grow bigger than the available memory, requiring a slow consumer to drop messages or terminate because of the memory overflow. Reactive Streams solves this problem by defining dynamic asynchronous pull-push with back pressure. Akka Streams implements the Reactive Streams standard using Akka which allows for seamless integration with both technologies.
Streams in Akka are built from blocks called stages or flows. These blocks can be nested and connected to each other, forming graphs. Graphs with single input and single output can be made runnable by connecting them to the source and sink. Graph definitions...