Warp-Speed z/OS to Kafka Messaging

Rapid zKafka Connector (RzK)

Connect your Mainframe directly to Kafka with a single, highly optimized API.

Any Language. Any Environment. Any Universe.

Maximizing ROI & Modernizing Core Systems

Business Value

Unmatched Performance & TCO Reduction

Stop burning expensive MIPS. By eliminating heavy Java clients, USS overhead, and MQ bridges, RzK’s native C/Assembler core processes significantly more messages per CPU second than traditional solutions.

Zero-Code-Change Integration

Seamlessly connect legacy systems into modern Event-Driven Architectures (Data Mesh). Produce and consume complete files via the RzK File Subsystem without altering existing programs.

Financial-Grade Reliability

Built-in support for Kafka Transactions ensures Exactly-once semantics. Critical financial data is never lost and never processed twice, completely mitigating integration risks.

Real-Time Data Streaming

Kafka’s core competencies (High Throughput, High Availability) finally unlocked directly on the mainframe without staging delays.

Integrating Kafka Data into Mainframe CICS Processes

RzK enables applications running under CICS to read (consume) data from Kafka or write (produce) data to it—without additional middleware.

Support for:

Transaction-based consume/produce operations.
Long-running tasks, short-running tasks and dialogs.
Full control over real-time data streams from a single central location.

Integrating Kafka Data into Mainframe DB2 Processes

DB2 ↔ Kafka: The mainframe application uses DB2 data and Kafka data simultaneously and updates either at the same time.

Also without any additional middleware.

Rapid zKafka Connector (RzK)

100% Native Architecture

The Integration Overview demonstrates our deep z/OS native approach. RzK utilizes a single 31-bit non-XPLINK core DLL with additional XPLINK DLLs. We provide dedicated Stubs (RZKSTUB) for LE programs, Call-Interfaces (RZKCALL) for non-LE, and even CICS TRUE integration.

Universal Language Support

Execute native CALL APIs from COBOL, PL/I, C/C++, Assembler, Fortran, Natural, REXX, or even Java (via JNI). No matter your environment (Batch, TSO, OMVS, CICS, IMS), RzK integrates seamlessly.

Event-Driven Mainframe in Action

Real-Time Monitoring (Grafana)

Stream operational metrics, SMF data, and subsystem performance indicators directly from z/OS into Kafka. Visualize your mainframe’s heartbeat in real-time dashboards.

Centralized Logging (Graylog)

Modernize your auditing. Use RzK to stream z/OS system logs and CICS transaction errors instantly to Kafka topics, seamlessly feeding into centralized log management systems.

Future-Proof Roadmap

zIIP Processor Offload

Officially approved and licensed by IBM. Automatically routes CPU workloads to zIIPs.

RzK File SubSystem

Coming Q4/2026. Allocate Kafka topics like standard datasets in your JCL.

RzK Connection Server

Coming Q4/2026. Centralized STC to reuse TCP/IP connections and reduce TLS handshakes.

RzK – Features & Capabilities

RzK – Universal Features

Standard z/OS CALL API: Enables integration into any z/OS application and programming language.
Supported Environments: Batch, TSO, OMVS/USS, CICS, IMS, DB2, Natural & RPC, LE & non-LE.
Kafka Transactions: Support for complex workflows through transactional security.
RzK Copy: Write messages directly into z/OS datasets or from z/OS datasets into the Kafka cluster—without any modifications to the mainframe application.
RzK File Subsystem: Read and write complete files between Kafka and the mainframe using standard file I/O, without program changes.
State-of-the-Art Security: Integrated SSL/TLS encryption and SASL authentication.
Broad Language Support: COBOL, C, C++, PL/I, Fortran, Natural, ASM, REXX, Java, Python, Perl, PHP.
Remote-Administration: Management of Kafka topics and groups directly from the mainframe.
Cross-Platform: The Rapid zKafka Connector is available for z/OS and Open Systems.

RzK – Ease of Installation and Operation

Easy Setup: Installing RzK does not require specialized mainframe expert knowledge.
Fast Start*: Ready for initial data exchange between z/OS and Kafka in just one hour.
Intuitive API: The interface is easy to understand and quick to implement.
Simplified Testing: Managing topics and groups directly from the mainframe accelerates the development of new features.
Background Management: RzK automatically manages complex connections to the Kafka cluster in the background.
Robustness: Automated recovery and error handling in the event of connection loss.
Transparency: Detailed error messages and status reports via the zKafka Connector API.
Debugging: Multiple diagnostic levels for efficient development and testing phases.

*) subject to pre-performed firewall settings

RzK – Flexibility and Efficiency

Direct Coupling: Direct connection between mainframe applications and Kafka clusters (no performance limiting detours via Java or USS calls).
Real-Time Processing: Kafka data is processed directly without time-consuming staging on the mainframe.
Resource-Efficient: Reduces inconsistencies, I/O operations, memory requirements, and CPU costs.
Low Overhead: Minimal complexity and reduced licensing costs, as no additional middleware is required.
zIIP Support (*): Significant cost savings by offloading CPU workloads to zIIP processors.
Optimization Options: Flexible trade-offs between performance and data consistency (e.g., via batching and compression).
Data Transformation: Asynchronous or synchronous transmission with built-in conversion between EBCDIC and UTF-8 formats.
Testing Tool: Easy-to-use Command Line Interface (RzK CLI) on z/OS.

(*) see Roadmap

Product Data Sheet / Flyer



RzK Product Flyer

Download

Evaluation Package

Download