rbreejen / oracle-logminer-flink-cdc-delta-lake

WIP: Demo Oracle/Postgres to Delta Lake/Hudi using Flink CDC and Debezium standalone

An exploration of Flink and change-data-capture (CDC).

docker-compose up --force-recreate
docker compose -f docker-compose.yaml -f docker-compose-oracle.yaml up --force-recreate
docker compose -f docker-compose.yaml -f docker-compose-postgres.yaml up --force-recreate --renew-anon-volumes
docker compose -f docker-compose-postgres-debezium.yaml up

In order to start the SQL Client run

docker compose run sql-client

Requirements

Some dependencies are needed for the Flink application to work properly which entail some explanation

• delta-flink1.16-bundle-1.0 - Fat jar with all required dependencies of Delta
• hadoop-aws (excl. aws-java-sdk-bundle), flink-s3-fs-hadoop (implicit aws-java-sdk-s3) - Necessary dependencies for the integration between S3A and Hudi/Delta without the big fat aws-java-sdk-bundle.
• hudi-flink1.16-bundle-0.13.1 - Package provided by Hudi developers, with all the necessary dependencies to work with the technology.
• flink-sql-connector-postgres-cdc-2.4.0 - Necessary dependencies for the integration between Postgres and Flink SQL CDC
• flink-sql-connector-oracle-cdc-2.5-SNAPSHOT - Necessary dependencies for the integration between Oracle 21 and Flink SQL CDC

How to create the fat jar delta-flink1.16-bundle-1.0

You can build the fat jar by using command maven clean package in directory delta-flink-bundle. The used pom.xml is mostly similar to the example pom.xml as specified by Delta.

Three different ways to implement CDC

Flink CDC & Flink SQL

Flink CDC & DataStream API

Add the application JAR using the Flink UI.

Using Debezium & Kafka Connect

Follow the instructions on how to use Debezium to monitor a Oracle database using Kafka Connect.

Show the Demo environment

The demo environment consists of the following components.

Minio (S3-compatible Storage)

• Web UI: http://localhost:9000 (user: minio, password: minio123)

Flink JobManager and TaskManager

• Queries are executed on a FLink cluster
• Web UI: http://localhost:8081