Kafka on Knative

The repository bootstraps a GKE cluster through Terraform and runs a Knative sink that consumes events from a Kafka producer.

Prerequisites

1. Terraform
2. GCP Cloud SDK
3. Kubectl

Setting Up the Kubernetes Cluster on GKE

1. Use an existing GCP service account or create a new service account with the following roles:
   • Kubernetes Engine Admin
   • Storage Admin (used for Terraform GCS backend)
   • Service Account User
   • Compute Viewer (used for retrieving the client certificate)
2. Move JSON credentials to ./account.json. This will be used for both the creation of the cluster and mutating the shared terraform state
3. Navigate to the ./terraform directory
4. Run terraform init
5. Modify the dev.auto.tfvars file
6. Run terraform apply

Setting Up kubectl

1. Run the following:

gcloud container clusters get-credentials <name>-cluster --zone <cluster_location> --project <project_id>