This is a Helm Chart for Scality's Zenko multi-cloud controller. To get started, you'll need a Kubernetes cluster, initialized with Helm. Check out minikube for a local or single-machine deployment. The below assumes you have cloned the Zenko repository and already cd into this directory.

Assuming you have an ingress controller running, run something like:

$ helm install --name zenko --set ingress.enabled=true \
--set ingress.hosts[0]=zenko.local \
--set cloudserver.endpoint=zenko.local zenko

You can disable integration with the Orbit management UI (enabled by default).

See values.yml for more options.

To disable Orbit, use:

--set cloudserver.orbit.enabled=false

If a custom Orbit endpoint is required, use:

--set cloudserver.orbit.endpoint=https://custom.orbit.endpoint
--set cloudserver.orbit.pushEndpoint=https://custom.push.orbit.endpoint

and if poll mode is needed, use:

--set cloudserver.orbit.mode=poll

Note: poll mode should only be used when long-lived websocket connections are not available. In poll mode:

• orbit.pushEndpoint setting will be ignored
• Interactive features such as Multicloud Browser, Metadata Search, CRR object status and CRR object retry will not be available
• Configuration updates may take up to a minute to propagate
• Statistics reports are sent less frequently

This chart can be configured to deploy a HorizontalPodAutoscaler for cloudserver. This is disabled by default.

To set up autoscaling based on CPU consumption, configure the amount of CPU a single cloudserver Pod requests. For example, to request one CPU to be allocated:

--set cloudserver.resources.requests.cpu=1

Next, enable autoscaling using:

--set cloudserver.autoscaling.enabled=true

This scales up (and down) between 1 and 16 replicas, with 80% CPU consumption as the per-Pod target.

These defaults can be modified using:

--set cloudserver.autoscaling.config.minReplicas=...
--set cloudserver.autoscaling.config.maxReplicas==...
--set cloudserver.autoscaling.config.targetCPUUtilizationPercentage=...

Prometheus is deployed as part of this stack. To access its built-in web UI, create an Ingress object. To do so automatically, use something like

--set prometheus.server.ingress.enabled=true --set prometheus.server.ingress.hosts[0]=prometheus.local

Grafana is deployed as a part of the Zenko chart. To access its built-in web UI, you can expose its service through any method that your platform supports. For example, if your Kubernetes cluster has an ingress controller running, you can expose the service by creating an Ingress object. To do so automatically during installation, use something like:

--set grafana.ingress.enabled=true --set grafana.ingress.hosts[0]=grafana.local

A test suite is shipped with the Zenko Chart to help you run validation tests on your deployment. Invoke this suite by running:

$ helm test $RELEASE_NAME --cleanup

where $RELEASE_NAME is replaced by the name you picked for the Helm release; zenko in the example above.

To upgrade an existing Zenko installation to the latest version or simply change deployment configuration, you will need to run the following commands. This assumes that your release name is zenko and that you have previously cloned the repository.

$ cd Zenko/kubernetes
$ git pull
$ helm upgrade zenko ./zenko

You may want to run an upgrade simulation, something highly recommended in production environments. For example:

$ helm upgrade zenko ./zenko --dry-run --debug

Note: In a production environment you may want to take neccessary additional steps to validate the upgrade before proceeding with the upgrade. Running the upgrade with the --dry-run flag will simulate and try to validate a compatible upgrade. Running with the --debug will output all the templated values and deployment configurations that will be installed. These are just basic validations but upgrade implications should be fully taken into account by you and/or your Kubernetes administrator.

Nightly releases are currently available under our private chart registry. They're designed for dev purposes only, so that you can get an early preview of what's about to be released.

They can be installed like the following:

# Add the zenko-dev repo to your helm setup
$ helm repo add zenko-dev https://registry.scality.com/chartrepo/zenko-dev
# Then you can install Zenko
$ helm install zenko-dev/zenko --version 1.2-nightly

To update your nightly release setup you follow this commands:

# To ensure your repo is up to date
$ helm repo update
# Then upgrade your release
$ helm upgrade my-release zenko-dev/zenko --version 1.2-nightly

For the nightly version scheme we use the Major and Minor version of Zenko (we use semver) suffixed by -nightly. So that it looks like the following ${MAJOR_VERSION}.${MINOR_VERSION}-nightly.

We also upload the same chart with a date suffix so that we can keep an history of every nightly.