Getting Started with the OperatorHub on OpenShift 4.x
The OperatorHub is a feature of OpenShift 4.x that enables browsing a catalog of open-source and vendor-provided operators. You can install operators from the operator catalog in the OpenShift administrative interface with a few simple steps.
You can install the operator via the Command-Line Interface (CLI), but the OperatorHub is an easy way to help you to manage the operator and apply upgrades. It also provides an easy way to create clusters.
The operator installation process is:
- Install the operator from the OperatorHub.
- Use a Custom Resource Definition (CRD) to create clusters from the Operator view in the OperatorHub.
- Create databases in Redis Enterprise.
Now the databases are exposed as Kubernetes services accessible by other application workloads in your K8s cluster.
Preparing the cluster
The Redis Enterprise node pods must run with certain privileges that are set in OpenShift using a Security Context Constraint that grants the pod various rights, such as the ability to change system limits or run as a particular user. At minimum, the security context constraint for the operator (scc.yaml) must be installed into the cluster as it is used by the OperatorHub installer. Without this constraint installed, the operator cannot create Redis Enterprise clusters.
The security context constraint for the operator needs to be installed only once and must not be deleted.
The constraint scc.yaml can be downloaded and installed by a cluster administrator with the commands:
curl -O https://raw.githubusercontent.com/RedisLabs/redis-enterprise-k8s-docs/master/openshift/scc.yaml oc apply -f scc.yaml
After the constraint is installed, the OperatorHub automatically uses the constraint for Redis Enterprise node pods.
Known Limitation - The automatic use of the security constraint is limited. The
Redis Enterprise must be named
rec for the constraint to be used automatically. This
limitation may be removed in the future. We recommended that you use the cluster name
rec when deploying with
If you require a different name, you must grant the SCC to the project
my-project) as in OpenShift 3.x:
oc adm policy add-scc-to-group redis-enterprise-scc system:serviceaccounts:my-project
Install the Operator
This procedure shows how you can install the Redis Enterprise Operator into a single project or namespace.
Step 1: Log in
The operator is deployed into a project. As such, you need an account with access to your project or the ability to create a new project.
Log in to your OpenShift account.
Either navigate to your project to set it as the default, or go to Projects and select Create Project.
In the dialog, enter the project name with the optional metadata and select Create.
Step 2: Navigate to database operators
- In the left menu, go to Catalog > OperatorHub.
- In the OperatorHub view, you see a list of categories on the left and a list of operators on the right. You can search for “Redis Enterprise” in the search dialog below the list of categories or select Database in the OperatorHub.
- Make sure that your project is shown in the “Project:” label at the top of the page.
Step 3: Find and select install
Either search for “Redis Enterprise” or scroll through the various database operators and select the Redis Enterprise Operator.
Then select the Install button.
Step 4: Create the operator subscription
When you use the OperatorHub, a deployed operator is maintained by the OpenShift Operator Lifecycle Management. The Redis Enterprise subscription that provides this is only for a single project. You cannot select All namespaces on the cluster.
Click Subscribe to start the subscription for your project.
The subscription is shown and the installation status is updated soon after. If you selected Manual approvals, the requires approval link is shown.
Step 5: Preview and approve install plan
Click requires approval to see the install plan that requires approval.
Click Preview Install Plan to preview the install plan.
The install plan shows the resources to be installed into your namespace.
Select the Approve button to install the operator.
After the approval, the status of the resources installed is updated.
Step 6: Return to the operator
Click redis-enterprise-operator-cert to select the operator.
The installed version and catalog source are shown along with other information.
- Click on the installed version for information about the current version and create clusters.
- Click on the catalog source to see the subscription.
Click on the installed version to see more information about the current version and create clusters.
Click on The catalog source to view the subscription.
After you install the Redis Enterprise operator into a project you can:
- Navigate to the installed version and use the Create New button to create a new cluster in your namespace.
- Click on Catalog > Installed Operators in the left menu to view the operators in your project.
- Click on Catalog > Operator Management to view your operator subscriptions.
Create a Redis Enterprise cluster
You can easily create a cluster with the operator in Installed Operators.
Step 1: Navigate to the installed operator
Click on Catalog > Installed Operators.
Find the Redis Enterprise Operator listing for your project.
Step 2: Access the operator API
Under Provided APIS for that listing, select RedisEnterpriseCluster.
Select Create Redis Enterprise Cluster to create a cluster.
The CRD for a Redis Enterprise cluster in YAML format is shown.
Step 3: Configure and create your cluster
Verify that you are creating the CRD in the correct project namespace. You can see this above the text edit and shown in the YAML located after
Edit the YAML definition.Note -The name of the cluster must be “rec” for deployments from within the OLM. For more information, see the
name:field description in Options for clusters.
Click Create to create and deploy the cluster.
The name of your cluster is shown in the list and you click on the cluster name to see more details about the cluster.
Create a database
Step 1: Access the Redis Enterprise UI
When the cluster is created, a login is generated using the
in the cluster CRD (e.g.,
email@example.com) and a password is generated by
the operator. The generated password is stored in a Kubernetes secret.
Go to Workloads > Secrets in the left menu:
Find the cluster name used in the CRD and click on the name to see your secret.
Scroll to the Data section to find the username and password.
Click copy to clipboard or Reveal Values to save the login information for later use.
Go to Workloads > Pods to see the pods associated with the cluster. These all have a number starting at “0” (e.g., rec-0) at the end of the name. Each of these pods has an instance of the Redis Enterprise UI running in it on port 8843.
To forward the port of one of these pods to your local machine, run:
kubectl port-forward rec-0 8443:8443
rec-0is replaced with of the pods associated with your Redis Entperise cluster.Note -
- rec-0 is one of your cluster pods. Consider running the port-forward command in the background.
- The Openshift UI provides tools for creating additional routing options, including external routes. These are covered in RedHat Openshift documentation.
In a browser, go to localhost:8443 to open the Redis Enterprise web UI:
Enter the username and password to log in to your cluster.
Step 2: Create your database
Continue with the instructions to create your database. After you create a database, the operator discovers the database and creates additional Kubernetes services for application workload access to the database.
Step 3: Inspect your database services
After you create your database in the Redis Enterprise web UI, the operator
detects the change and creates Kubernetes services that expose the database. The
databases are named according to the database name. For example, if you
called your database “
test”, kubectl shows these services:
% kubectl get services NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE rec ClusterIP None <none> 9443/TCP,8001/TCP,8070/TCP 43m rec-ui ClusterIP 172.30.16.113 <none> 8443/TCP 43m test ClusterIP 172.30.212.143 <none> 13818/TCP 20s test-headless ClusterIP None <none> 13818/TCP 20s
The operator creates two services, shown here as “
test” and “
The service with the “-headless” suffix is a Kubernetes headless service and does not provide load balancing.