⚙️ CrateDB Kubernetes Operator

The CrateDB Kubernetes Operator provides convenient way to run CrateDB clusters inside Kubernetes. It is built on top of the Kopf: Kubernetes Operators Framework.

🗒️ Contents

• 🤹 Usage
• 🎉 Features
• 💽 Installation
• 💻 Development

🤹 Usage

A minimal custom resource for a 3 node CrateDB cluster may look like this:

dev-cluster.yaml:

apiVersion: cloud.crate.io/v1
kind: CrateDB
metadata:
  name: my-cluster
  namespace: dev
spec:
  cluster:
    imageRegistry: crate
    name: crate-dev
    version: 5.7.3
  nodes:
    data:
    - name: hot
      replicas: 3
      resources:
        limits:
          cpu: 4
          memory: 4Gi
        disk:
          count: 1
          size: 128GiB
          storageClass: default
        heapRatio: 0.5

$ kubectl --namespace dev create -f dev-cluster.yaml
...

$ kubectl --namespace dev get cratedbs
NAMESPACE   NAME         AGE
dev         my-cluster   36s

Please note that the minimum version of CrateDB that the operator supports is 4.5. Previous versions might work, but the operator will not attempt to set a license.

🎉 Features

• "all equal nodes" cluster setup
• "master + data nodes" cluster setup
• safe scaling of clusters
• safe rolling version upgrades for clusters
• SSL for HTTP and PG connections via Let's Encrypt certificate
• custom node settings
• custom cluster settings
• custom storage classes
• region/zone awareness for AWS and Azure

💽 Installation

Installation with Helm

To be able to deploy the custom resource CrateDB to a Kubernetes cluster, the API needs to be extended with a Custom Resource Definition (CRD). It can be installed separately by installing the CRD Helm Chart or as a dependency of the Operator Helm Chart.

helm repo add crate-operator https://crate.github.io/crate-operator
helm install crate-operator crate-operator/crate-operator

To override the environment variables from values.yaml, please refer to the configuration documentation.

Installation with kubectl

To be able to deploy the custom resource CrateDB to a Kubernetes cluster, the API needs to be extended with a Custom Resource Definition (CRD). The CRD for CrateDB can be found in the deploy/ folder and can be applied (assuming sufficient privileges).

$ kubectl apply -f deploy/crd.yaml
customresourcedefinition.apiextensions.k8s.io/cratedbs.cloud.crate.io created

Once the CRD is installed, the operator itself can be deployed using a Deployment in the crate-operator namespace.

$ kubectl create namespace crate-operator
...
$ kubectl create -f deploy/rbac.yaml
...
$ kubectl create -f deploy/deployment.yaml
...

Please refer to the configuration documentation for further details.

💻 Development

Please refer to the Working on the operator section of the documentation.