Neo4j-Helm

This repository contains a Helm chart that starts Neo4j >= 4.0 Enterprise Edition clusters in Kubernetes.

Full Documentation can be found here

Quick Start

Check the releases page and copy the URL of the tgz package. Make sure to note the correct version of Neo4j.

Standalone (single server)

$ helm install mygraph RELEASE_URL --set core.standalone=true --set acceptLicenseAgreement=yes --set neo4jPassword=mySecretPassword

Casual Cluster (3 core, 0 read replicas)

$ helm install mygraph RELEASE_URL --set acceptLicenseAgreement=yes --set neo4jPassword=mySecretPassword

When you're done: helm uninstall mygraph.

Documentation

The User Guide contains all the documentation for this helm chart.

The Neo4j Community Site is a great place to go for discussion and questions about Neo4j & Kubernetes.

Additional instructions, general documentation, and operational facets are covered in the following articles:

• Architectural Documentation describing how the helm chart is put together
• External exposure of Neo4j clusters on Kubernetes - how to use tools like Neo4j Browser and cypher-shell from clients originating outside of Kubernetes
• Neo4j Considerations in Orchestration Environments which covers how the smart-client routing protocol that Neo4j uses interacts with Kubernetes networking. Make sure to read this if you are trying to expose the Neo4j database outside of Kubernetes
• How to Backup Neo4j Running in Kubernetes
• How to Restore Neo4j Backups on Kubernetes

Helm Testing

This chart contains a standard set of helm chart tests, which can be run after a deploy is ready, like this:

helm test mygraph

Local Testing & Development

Template Expansion

To see what helm will actually deploy based on the templates:

helm template --name-template tester --set acceptLicenseAgreement=yes --set neo4jPassword=mySecretPassword . > expanded.yaml

Full-Cycle Test

The following mini-script will provision a test cluster, monitor it for rollout, test it, report test results, and teardown / destroy PVCs.

Provision K8S Cluster

Please use the tools/test/provision-k8s.sh, and customize your Google Cloud project ID.

Standalone

Standalone forms faster so we can manually lower the liveness/readiness timeouts.

export NAME=a
export NAMESPACE=default
helm install $NAME . -f deployment-scenarios/ci/standalone.yaml && \
kubectl rollout status --namespace $NAMESPACE StatefulSet/$NAME-neo4j-core --watch && \
helm test $NAME --logs | tee testlog.txt
helm uninstall $NAME
sleep 20
for idx in 0 1 2 ; do
  kubectl delete pvc datadir-$NAME-neo4j-core-$idx ;
done

Causal Cluster

export NAME=a
export NAMESPACE=default
helm install $NAME . -f deployment-scenarios/ci/cluster.yaml && \
kubectl rollout status --namespace $NAMESPACE StatefulSet/$NAME-neo4j-core --watch && \
helm test $NAME --logs | tee testlog.txt
helm uninstall $NAME
sleep 20
for idx in 0 1 2 ; do
  kubectl delete pvc datadir-$NAME-neo4j-core-$idx ;
done

Internal Tooling

This repo contains internal tooling containers for backup, restore, and test of the helm chart.

Building the Containers

If you want to push your own docker containers, make sure that the registry in the Makefile is set to somewhere you have permissions on.

cd tools
make docker_build
make docker_push