Introduction

Kafka-conn is a Kafka Connector provisioning as a code following Gitops concept by using Gitlab Pipeline.

Deployment as Code (GitOps style)

Installtion

Prerequisite

• Docker
• Kafka Confluent by Kubernetes Helm Charts: https://docs.confluent.io/5.0.0/installation/installing_cp/cp-helm-charts/docs/index.html
• Gitlab Pipeline

Run

$ docker build . -t kafka-conn:v1
$ docker run kafka-conn:v1 kafka-conn --help
Usage: /app/kafka-conn [--help -h] 
Eg: 
         /app/kafka-conn -h --help 
         /app/kafka-conn validate env/dev
         /app/kafka-conn apply env/dev
         /app/kafka-conn sync env/dev

Instructions for Use

Folder structure

$ tree env
├── env
│   ├── dev
│   │   ├── __sink-minio-kong
│   │   └── __source-orders-dev
│   └── stag
│       └── __sink-echo-s3

• Each file with prefix __<sink/source>-connectorName will be configured via REST API to Kafka-Connector. To define which Kafka-connector is used via OS variable $KAFKA_CONNECT_URL
• Let's check .gitlab-ci.yml for more info.

Gitlab Pipeline

• To make sure all PR/Commit for creating a connector is validated, I proposed this pipeline to validate, apply and clean up all connector in the kafka.

stages:
- validate # All jobs in this stage will check all connector syntax, json validation...etc
- apply # This will be run kafka-conn apply to the connector.
- cleanup # This job cleans up all unused and undefined in env/* folder.

Acknowledgements

• https://docs.gitlab.com/ee/ci/pipelines/
• https://docs.confluent.io/platform/current/connect/references/restapi.html