Oracle TxEventQ and Apache Kafka Integration: Two Powerful Paths to Modern Event Streaming

Introduction

Oracle Transactional Event Queues (TxEventQ) brings event streaming natively inside the Oracle AI Database. Built on top of Advanced Queuing (AQ) but enhanced for scale, partitioning, and pub-sub semantics, TxEventQ eliminates the need for a separate broker infrastructure in many architectures.

To make integration seamless for Kafka-based systems, Oracle offers two key options:

1. Kafka Java Client for TxEventQ (OKafka) – a brokerless replacement path.

2. Kafka Connect Source and Sink Connectors for TxEventQ – a bridge between Oracle AI Database and existing Kafka clusters.

Each approach serves a different need—either simplifying your stack or enabling hybrid interoperability.

Option A: Kafka Java Client for TxEventQ (OKafka)

What It Is

OKafka is a Kafka-compatible Java client that replaces the external Kafka broker with the Oracle AI Database itself. It implements familiar Kafka APIs such as Producer, Consumer, and Admin, but routes messages through JDBC and AQ to TxEventQ topics inside the database.
This allows existing Kafka applications to run with minimal changes—mostly configuration adjustments.

How It Works

In the OKafka path, Kafka topics map directly to TxEventQ queues.
Partitions correspond to database-level partitions, and offsets are maintained transactionally.
Every enqueue (produce) and dequeue (consume) operation participates in the same Oracle transaction, guaranteeing atomicity and exactly-once semantics.

Your producer and consumer code stays nearly identical; you only point bootstrap.servers to the Oracle listener, configure JDBC or wallet-based credentials, and use uppercase topic names.
The Oracle AI Database acts as both the event store and message broker, maintaining persistence, ordering, and recovery automatically.

Setup Basics

1. Ensure you are using Oracle AI Database 21c or later, where TxEventQ is available.

2. Create or grant a user with AQ privileges (DBMS_AQADM, DBMS_AQ, etc.).

3. Set an appropriate STREAMS_POOL_SIZE and ensure the database listener is reachable.

4. Add OKafka and Oracle JDBC/AQ libraries to your application classpath.

5. Create a TxEventQ topic via DBMS_AQADM.CREATE_DATABASE_KAFKA_TOPIC and reference it from your producer.

Example Usage

Properties props = new Properties();
    props.put("bootstrap.servers", "dbhost:1521");
    props.put("oracle.service.name", "ORCLPDB1");
    props.put("security.protocol", "PLAINTEXT");
    props.put("oracle.net.tns_admin", "/path/to/tns");
    
    AdminClient admin = AdminClient.create(props);
    NewTopic topic = new NewTopic("MY_TOPIC", 3, (short)1);
    admin.createTopics(Collections.singleton(topic)).all().get();
    
    KafkaProducer<String,String> producer =
        new KafkaProducer<>(props, new StringSerializer(), new StringSerializer());
    producer.send(new ProducerRecord<>("MY_TOPIC", "key", "value")).get();
    producer.close();
    admin.close();