Lesson 4 of 5
Integration patterns
9 min
Listen to this lesson
A technology says how bytes travel. A pattern says how systems are arranged. Three patterns cover most SAP landscapes: point-to-point, mediated (hub-and-spoke) and event-driven. Recognizing the pattern tells you who depends on whom, where to test, and what breaks when one system changes.
Point-to-point
In a point-to-point pattern, each sender connects directly to each receiver. It is quick to build for the first few interfaces: the shop-floor system in Detroit calls a BAPI in SAP directly. The problem is growth. With many systems, every new connection adds another custom link, and nobody has a single view of the traffic. A change to one system can break links that nobody remembers exist.
Mediated (hub-and-spoke)
In a mediated pattern, systems connect to a central platform, which transforms and routes messages. SAP Process Orchestration and SAP Integration Suite are mediators. The benefits are central monitoring, reusable mappings and one place to change routing. The cost is that the hub becomes critical: when it is down, everything waits, and every change to it needs testing across many interfaces.
Event-driven
In an event-driven pattern, a system publishes an event, such as "sales order created", to a broker. Any system that is interested subscribes. The publisher does not know or care who listens. This decouples systems well: Contoso can add a new subscriber, for example a fleet analytics service, without touching SAP S/4HANA. Events are usually thin notifications; a subscriber that needs the full order calls an API to fetch it.
| Pattern | Shape | Strength | Weakness | Test focus |
|---|---|---|---|---|
| Point-to-point | Direct link per pair of systems | Simple and fast for a few interfaces | Grows into a tangle; no central view | Each link separately; hidden dependencies |
| Mediated | All links through a hub | Central monitoring, mapping and routing | Hub is a single critical component | Mappings, routing rules, hub capacity |
| Event-driven | Publish to a broker; subscribers react | Loose coupling; easy to add consumers | Harder to trace end to end; ordering and duplicates | Event delivery, subscriber behavior, replay |
Patterns by use case
SAP's Integration Solution Advisory Methodology (ISA-M) describes integration in terms of integration styles, such as process integration and data integration, refined into use case patterns. For the process integration style, SAP Learning lists patterns including A2A (application-to-application), B2B (business-to-business), B2G (business-to-government) and master data integration. These names help teams talk about the business purpose of an interface before arguing about technology.
| Use case pattern | Contoso example | Typical technology |
|---|---|---|
| A2A | SAP S/4HANA to the warehouse and shop-floor systems | IDoc, RFC, OData through the platform |
| B2B | Orders from Great Lakes Fleet Services; POs to Lakeside Components | EDI translated to IDocs via the platform |
| B2G | Electronic tax filings to authorities, relevant for the planned Mexican acquisition | Country-specific SAP solutions and APIs |
| Master data integration | Product data replicated to the e-commerce catalog | IDoc or APIs, increasingly events |
ISA-M process integration use case patterns with Contoso examples.
What would you do?
An architect proposes connecting a new fleet analytics service directly to the SAP S/4HANA database tables to read delivery data. What is the best response?
- Point-to-point is simple at first and tangled at scale.
- Mediated integration centralizes mapping, routing and monitoring, but makes the hub critical.
- Event-driven integration decouples publishers from subscribers; trace and duplicate handling need care.
- ISA-M names use case patterns such as A2A, B2B, B2G and master data integration.
Check your understanding · not graded
Lesson check
1. What is the main weakness of point-to-point integration as a landscape grows?
2. In an event-driven pattern, what does the publisher know about its subscribers?
3. Orders from Great Lakes Fleet Services over EDI are which ISA-M use case pattern?