liveBook · Manning

Integration mode

As explained in Clean Applications with Hexagonal Architecture

An integration mode is a way for work or information to cross an architectural boundary, distinguished by waiting and responsibility for continuing work.

An integration mode describes how work or information crosses an architectural boundary: who waits and who is responsible for continuing the work. section 10.3

Choose among three modes for each interaction:

  • Synchronous call: the caller waits for an answer before continuing. section 10.3
  • Event: the publisher hands off a fact and continues; receivers may react later. section 10.3
  • External orchestration: a coordinator tracks a sequence across steps and can hold it durably for recovery. section 10.4

Choose a synchronous call when the next step needs an answer, an event when others only need to be informed, and external orchestration when a process must span systems or survive failure. Internal orchestration within one process is the handler’s responsibility, not an integration mode. section 10.4 section 10.7

WriteFlow interactions

A CMS waits for a decision to reopen an article; subscribers can learn later that an article was approved; a coordinator can track a longer editorial pipeline. One system may use different modes for different interactions. section 10.0 section 10.3 section 10.4

Figure 10.1 The decisions for one interaction produce one contract-map entry. Here, the CMS’s request is inbound, synchronous because the CMS needs WriteFlow’s decision before continuing, and governed by a contract WriteFlow owns.

Use the contract-map figure to see how direction, waiting, and contract ownership determine an entry for the CMS request. section 10.0

How Clean Applications with Hexagonal Architecture covers it

How deep it goes
Distinguishes three modes and gives timing and process-lifetime criteria, while deferring coordinator-tool selection.
Examples
Uses a CMS reopen request, approval notifications, an editorial pipeline, and a mixed publish path.
What it assumes
Assumes a boundary between a domain core and external systems, and distinguishes requests from facts.
What you can do after
Apply a mode per interaction based on answer timing and process durability needs.
How it connects
Connects synchronous calls, events, external orchestration, requests, facts, and handlers.

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Clean Applications with Hexagonal Architecture. The section links point to the parts of the book it draws on.