Note: This project is archived. The recommended way for running Debezium as well as Apache Kafka, Kafka Connect on ZooKeeper on Kubernetes and distributions such as OpenShift is to use the custom resource definitions (CRDs) and operators provided by the Strimzi project.
This project builds the kubernetes manifest files for running Zookeeper (standalone at the moment), Apache Kafka, Kafka Connect with the Debezium MySQL connector and a MySQL 5.6 database on Kubernetes. The project uses the awesome fabric8-maven-plugin for automatically generating the manifest in both json and yaml formats.
To build, go to the root of this project and run:
mvn clean install
Then you can look in the target/classes folder of each subproject and see the kubernees.json and kubernetes.yml files. Each file specifies a Kubernetes ReplicationController (now named ReplicaSet in Kubernetes 1.2+) and a Kubernetes Service.
If you're logged in to a Kubernees cluster, you can deploy the json/yml with maven:
mvn fabric8:apply
This will read your ~/.kube/config file and deploy it into the namespace which is in your current context. Please see the fabric8-maven-plugin documentation for more details and fine grained control..
If you don't want to use Maven to deploy, you can use the kubectl tool directly. For example:
kubectl create -f zk-standalone/target/classes/kubernetes.yml