Reactivating enterprise integration

Description

As software grows more and more interconnected, and with several generations of software having to interoperate, a new take on the integration of systems is needed. Ad hoc, unversioned, and unreplicated scripts just won’t suffice, and enterprise service buses have known stability, reliability, performance, and scalability problems.

Viktor Klang offers a new take on enterprise integration patterns that builds on top of the Reactive Streams standard, an orchestration layer where transformations are standalone, composable, and reusable and—most importantly—use asynchronous flow-control (backpressure) to maintain predictable, stable behavior over time. Viktor also explains how one-off workloads relate to continuous and batch workloads and how they can be addressed by that very same orchestration layer. He concludes with a discussion of how a design like this achieves resilience, scalability, and ultimately, responsiveness.