Skip to content

Latest commit

 

History

History
87 lines (56 loc) · 3.8 KB

README.md

File metadata and controls

87 lines (56 loc) · 3.8 KB
Raw

Apache Kafka on K8s using Helm

About this repo

This repository is deploying Apache Kafka and other third party integrations on Kubernetes using a Helm chart.

What's going to be deployed

  • Namespace: a new namespace called kafka that will have all the deployed resources.

  • RBAC: a new ServiceAccount, ClusterRole, and ClusterRoleBinding to manage access between different resources.

  • Kafka-broker: the Apache Kafka broker will be deployed as a StatefulSet using the latest kafka docker image. A headless service will be used to expose it.

  • Kafka configuration: all Kafka settings can be controlled from the broker-config ConfigMap file, including the init script "init.sh" and both: server.properties and log4j.properties.

  • Zoo Keeper: will be dployed as a StatefulSet, along with ClusterIP servcie (expose both 2888 and 3888 ports), headless service (expose 2181 port), and a ConfigMap settings file.

  • Persistent Volumes: tow PersistentVolume and PersistentVolumeClaim will be provisioned to persist the data within for both kafka-broker and zoo keeper.

  • AKHQ: Kafka GUI for Apache Kafka to manage topics, topics data, consumers group, schema registry, connect and more...

  • ZooNavigator: it's a web-based ZooKeeper UI and editor/browser with many features.

  • kafka-cli: a kafka-based k8s pod to interact with both kafka-broker and zookeeper programatically.

How to deploy the chart:

  • In order to deploy the above resourtce, you will need to install Helm, a package manager for K8s. More info and installation guide can be found below: https://helm.sh/docs/intro/install/

  • Once Helm is installed,m the above chart can be deployed using the below commands:

git clone git@github.com:meniem/kafka-helm.git
helm upgrade --install kafka-chart ./helm

How to deploy the Kafka cluster locally:

  • You can create a local k8s environment to test the resopurces using Minikube, which runs a single-node Kubernetes cluster on your personal computer. To install and start a new minikube cluster, please follwo the steps below:

https://kubernetes.io/docs/tasks/tools/

https://minikube.sigs.k8s.io/docs/start/

  • Once the local k8s lcouster is up and running, you casn follwo the same steps above to apply the helm chart.

Exposing the web-based tools locally:

  • Both AKHQ and ZooNavigator tools are deployed using NodePort service type to easily expose them locally.

  • To expsoe both services using Minikube, run the below commands:

minikube service --url akhq -n kafka
minikube service --url zoonavigator -n kafka

Note: for ZooNavigator, when asking for the connection string, type: zookeeper-svc.kafka.svc.cluster.local:2181 as this is the FQDN of the service on the cluster.

Interacting with the cluster from the CLI:

  • The kafka-cli pod is used to interact with the cluster from CLI, you can open an SSH session with the pod and test the basic kafka opreations as shown below:
# ssh into the pod:
kubectl exec -it kafka-cli -n kafka -- bash
# create a new topic
./bin/kafka-topics.sh --create --zookeeper zookeeper-svc:2181 --replication-factor 1 --partitions 1 --topic testtopic
>Created topic "testtopic".
# push a created topic:
./bin/kafka-console-producer.sh --broker-list kafka-broker:9092 --topic testtopic
>Message 1 in testtopic
# Open another ssh session to access the consumer and check the pushed message:
kubectl exec -it kafka-cli -n kafka -- bash
./bin/kafka-console-consumer.sh --bootstrap-server kafka-broker:9092 --topic testtopic --partition 0 --from-beginning
>Message 1 in testtopic