Skip to main content

Multi-Protocol Gateway: Overview

The Multi-Protocol Gateway (MPGW) is the service you'll build most often. It accepts requests on one or more protocols, runs them through a processing policy, and forwards them to a backend — optionally over a different protocol. "Multi-protocol" is the whole point: front and back can differ.

Anatomy​

An MPGW is configured from a few key pieces:

  • One or more front-side handlers — how clients reach it (protocol + address + port). One MPGW can listen on several at once (e.g. HTTP and MQ).
  • A processing policy — what it does with each message (the rules and actions). Covered in the next two lessons.
  • A backend type — where it sends the result:
    • Static backend — always the same fixed backend URL.
    • Dynamic backend — the destination is decided at runtime by the policy (e.g. route based on the URL or a header).
  • Request/response type — how to treat the message body: XML, JSON, SOAP, Non-XML (pass-through/binary), or Pass-Thru.

The request/response lifecycle​

Every message flows through the same lifecycle. An MPGW policy has a request rule (client → backend) and usually a response rule (backend → client), plus optional error rules.

If anything throws an error, processing jumps to the error rule (Module 5).

A mental model​

Hold these three questions in your head whenever you design an MPGW:

  1. How do clients reach it? → front-side handler(s)
  2. What happens to the message? → processing policy (request/response rules)
  3. Where does it go? → static or dynamic backend

The next lessons take each in turn, starting with handlers.

Knowledge Check

What does "multi-protocol" actually mean for an MPGW?

Which backend type lets the processing policy decide the destination at runtime?