Apache Kafka is a proven high-throughput event streaming platform that works as a universal message broker between your SAP systems and the broader application landscape. YUNIT SL designs and implements Kafka-based integrations deployed in Docker containers on hyperscalers such as AWS.
Two Primary Integration Scenarios
1. Kafka as Event Consumer (SAP produces events)
SAP backend systems generate business events — order placed, invoice created, shipment dispatched — and publish them to Kafka topics via REST adapter interfaces or RAP applications on SAP BTP ABAP Environment. Downstream systems (commerce frontends, analytics, third-party fulfilment) consume these events asynchronously.
- SAP Integration Suite processes outbound messages and converts them to Kafka-compatible formats
- IDocs used for structured data replication to Kafka topics
- RFC adapter calls for real-time SAP function module invocations
2. Kafka as Event Producer (SAP consumes events)
External systems — commerce platforms, marketplaces, IoT, customer portals — publish events to Kafka. The SAP Integration Suite picks up these messages, transforms them, and triggers corresponding SAP processes (order creation, customer update, stock reservation).
- SAP Integration Suite acts as Kafka consumer, transforming messages to SAP-compatible formats
- RAP applications on SAP BTP ABAP Environment handle complex business logic before posting to S/4 HANA
- Error queues and dead-letter handling for resilient processing
Deployment Architecture
Docker containers deploy the Kafka client to hyperscalers (AWS, Azure, GCP). This means:
- Infrastructure-as-code for repeatable deployments
- Independent scaling of producers and consumers
- No SAP BTP lock-in — Kafka connects to any system
- Blue-green deployment compatible with existing CI/CD pipelines