The diagram below illustrates the high-level architecture of the OneConnect solution. It depicts SAP ECC or S/4HANA as the data producer, with multiple downstream systems—such as ClickHouse, Snowflake, and Databricks—acting as subscribers. Other destinations can also be supported.
Responsible for modeling SAP entities based on SAP tables, CDS views, or ABAP programs.
Replication Methods:
Table Format: Each SAP table or CDS view is transformed into a Kafka topic, preserving primary keys—similar to Change Data Capture (CDC).
Document Format: Each SAP document is converted into a Kafka topic with a nested, hierarchical structure. This structure resembles an IDOC, but instead of segments, it contains SAP tables or CDS views.
Data Delivery:
Pushes both data and metadata to the Smart Gateway, either in batch mode or via real-time SAP event triggers (e.g., BOR Events, RAP Events).
A Kubernetes-based component that ingests data and metadata from SAP.
Translates the incoming content into Confluent-compatible schemas and Kafka topics.
Serializes messages into Avro format for efficient transport and schema enforcement.
Enterprise-grade connectors designed to streamline Kafka integrations with modern analytics platforms.
Supported Targets: Databricks, Snowflake, ClickHouse, and Apache Iceberg.
Key Features:
Rapid deployment of streaming pipelines
Automated schema evolution
Idempotent writes for consistent data ingestion
Secure, high-performance data transfers
These connectors enable accelerated data modernization and deliver measurable business value—often within hours of deployment.