OPAL Helm Chart for Kubernetes
OPAL is an administration layer for Open Policy Agent (OPA), detecting changes to both policy and policy data in realtime and pushing live updates to your agents.
OPAL brings open-policy up to the speed needed by live applications. As your application state changes (whether it's via your APIs, DBs, git, S3 or 3rd-party SaaS services), OPAL will make sure your services are always in sync with the authorization data and policy they need (and only those they need).
OPAL Helm chart could be installed only with Helm 3. The chart is published to public Helm repository, hosted on GitHub itself. It's recommended to install OPAL into a dedicated namespace.
Add Helm repository
helm repo add permitio https://permitio.github.io/opal-helm-chart
helm repo update
Install the latest version
helm install --create-namespace -n opal-ns opal permitio/opal
Search for all available versions
helm search repo opal --versions
Deploy OPAL to your Kubernetes cluster
Install specific version (with default configuration):
helm install --create-namespace -n opal-ns --version x.x.x opal permitio/opal
Install specific version (with custom configuration provided as YAML):
helm install -f myvalues.yaml --create-namespace -n opal-ns --version x.x.x opal permitio/opal
myvalues.yaml must conform to the json schema.
OPAL Client should populate embedded OPA instance with polices and data from configured Git repository. To validate it - one could create port-forwarding to OPAL client Pod. Port 8181 is the embedded OPA agent.
kubectl port-forward -n opal-ns service/opal-client 8181:8181
Then, open http://localhost:8181/v1/data/ in your browser to check OPA data document state.
This is not a comprehensive list, but includes the main variables you have to think about
|Git repository holding policy code (& optionally policy data) to be tracked by OPAL|
|Data sources to be published to clients (and their managed OPAs)|
|Static list of data source entries (See OPAL Docs)|
|URL to dynamically fetch data sources entries from (See OPAL Docs)|
|Backend for broadcasting updates across multiple opal-server processes (necessary if either |
|Count of gunicorn workers (/processes) per opal-server replica|
|opal-server's deployment replica count|
|Extra configuration for opal-server (see OPAL Docs)|
|Extra configuration for opal-server OPAL Docs|
Note: If you leave
server.dataConfigSources with no entries - The chart would automatically set
OPAL_DATA_UPDATER_ENABLED: False in
client.extraEnv so client won't report an unhealthy state.