Learning Path
Get started using Kubernetes Operators
- Overview
- Intro to Kubernetes Operators
- Deep dive into Kubernetes operators
- Develop and deploy a Level I JanusGraph Operator using BerkeleyDB
- Develop and deploy a Level I JanusGraph operator using Apache Cassandra
- The Operator Cookbook: How to make an operator
- Seamlessly upgrade a JanusGraph operator
- Certify a JanusGraph image and operator for publishing to the Red Hat Marketplace
- Summary
To follow the steps in this learning path for developing Golang-based operators, you need to install the following software on your machine. Environment-specific installation instructions for installing Operator SDK and the oc or kubectl CLIs and for provisioning a cluster are below.
Prerequisites for installing Operator SDK on a macOS
Note: The tutorials in this series work with Go v1.15 and Operator SDK v1.5.0.
To develop Golang-based operators on macOS, you need to install the following:
- Homebrew
- Go 1.10+
- Kustomize
- kubectl
- Docker v17.03+
- OpenShift CLI (If you plan to deploy to OpenShift cluster) oc
- To use an OpenShift cluster, we recommend using version 4.6+.
Additionally, you need:
- Access to a Kubernetes v1.11.3+ cluster (v1.16.0+ if using apiextensions.k8s.io/v1 CRDs). See minikube or CodeReady Containers to access a free cluster.
- Cluster admin permission. See how to grant yourself cluster-admin privileges or be logged in as admin.
- Access to a container registry such as Quay.io or DockerHub
Prerequisites for installing opeartor-sdk on Linux and Windows
To develop Golang-based operators on Linux and Windows, you need to install the following:
- dep v0.5.0+
- Git
- Go v1.10+
- Docker v17.03+
- OpenShift CLI (oc) v4.1+ installed
- Kustomize
- OpenShift CLI (If you plan to deploy to OpenShift Cluster) oc
- To use an OpenShift cluster, we recommend you use version 4.6+
Additionally, you need:
- Access to a Kubernetes v1.11.3+ cluster (v1.16.0+ if using apiextensions.k8s.io/v1 CRDs). See minikube or CodeReady Containers to access a cluster for free.
- Access to a container registry such as Quay.io or DockerHub
Steps
1. Install Operator SDK
Install Operator SDK (version 1.5+) with Go (version 1.15.5) for macOS
Normally, to install the Operator SDK with Homebrew, you would run brew install operator-sdk. This installs the latest version of the Operator SDK with the latest version of Go. However, the Operator SDK (as of v1.5) does not work with the latest version of Go (v1.16+). To follow these tutorials, you need to explicitly install the versions that are used in these tutorials: Go v1.15 and Operator SDK v1.5.0
- Install Homebrew if you don't have it.
- If you have a version of Go later than v1.15.5 installed, uninstall it. Then run
brew install go@1.15.5to install the required version of Go. - Run
brew install operator-sdk@1.5to install the version of the Operator SDK used for these tutorials. It runs properly with Go v1.15.5.
Install Operator SDK (version 1.5+) for Linux or Windows
Make sure to install Go v1.15.5
- For Linux or Windows, install the Operator SDK (version 1.0+) from the GitHub release. Note that commands have changed with the 1.0 release.
Test your environment for Operator SDK
Run the following command in the terminal of your choice:
$ operator-sdk versionYou should see output like this:
operator-sdk version: "v1.5.0", commit: "1abf57985b43bf6a59dcd18147b3c574fa57d3f6", kubernetes version: "v1.19.4", go version: "go1.15.5", GOOS: "darwin", GOARCH: "amd64"Ensure kustomize is installed. Install it via brew for macOS:
$ brew install kustomize
$ kustomize version
You should see output like this:
{Version:kustomize/v3.9.1 GitCommit:7439f1809e5ccd4677ed52be7f98f2ad75122a93 BuildDate:2020-12-30T01:08:17+00:00 GoOs:darwin GoArch:amd64}
2. Install oc or kubectl cli
If you plan to use an OpenShift cluster, install the OpenShift CLI (oc) using instructions in the OpenShift docs.
Otherwise, install kubectl from the Kubernetes docs.
Alternatively, you can install the oc CLI through the OpenShift web console as this article describes.
Test your CLI by issuing the following command to see the version of your CLI:
$ oc version
Client Version: openshift-clients-4.5.0-202006231303.p0-18-g6082e941e
Kubernetes Version: v1.19.2
If you plan to use kubectl instead of oc use:
$ kubectl version
Client Version: version.Info{Major:"1", Minor:"20", GitVersion:"v1.20.2", GitCommit:"faecb196815e248d3ecfb03c680a4507229c2a56", GitTreeState:"clean", BuildDate:"2021-01-14T05:15:04Z", GoVersion:"go1.15.6", Compiler:"gc", Platform:"darwin/amd64"}
Server Version: version.Info{Major:"1", Minor:"18+", GitVersion:"v1.18.3+e574db2", GitCommit:"e574db2", GitTreeState:"clean", BuildDate:"2021-01-30T06:33:00Z", GoVersion:"go1.13.15", Compiler:"gc", Platform:"linux/amd64"}
3. Provision an OpenShift cluster
(If you haven't already) provision an OpenShift cluster by going to https://cloud.ibm.com/ and clicking Red Hat OpenShift on IBM Cloud tile. From there, you can select an OpenShift cluster of your choosing.
We recommend you use an OpenShift cluster of version 4.6 or higher.
After the cluster finishes provisioning, you should get a message that the Worker nodes are 100% normal, and the Master status is ready, as the picture below shows.
Provisioning an OpenShift cluster takes some time, so do so now if you don't have one already. Skip down to the prepare your OpenShift Cluster step to see how to create an OpenShift cluster on IBM Cloud.
Ensure OpenShift Lifecycle Manager (OLM) is updated
To make sure your OpenShift Lifecycle Manager is
up to date and running properly, run the operator-sdk olm status command:
operator-sdk olm status
INFO[0003] Fetching CRDs for version "0.16.1"
INFO[0003] Using locally stored resource manifests
INFO[0005] Successfully got OLM status for version "0.16.1"
NAME NAMESPACE KIND STATUS
operators.operators.coreos.com CustomResourceDefinition Installed
operatorgroups.operators.coreos.com CustomResourceDefinition Installed
installplans.operators.coreos.com CustomResourceDefinition Installed
clusterserviceversions.operators.coreos.com CustomResourceDefinition Installed
olm-operator olm Deployment Installed
subscriptions.operators.coreos.com CustomResourceDefinition Installed
olm-operator-binding-olm ClusterRoleBinding Installed
operatorhubio-catalog olm CatalogSource Installed
olm-operators olm OperatorGroup Installed
aggregate-olm-view ClusterRole Installed
catalog-operator olm Deployment Installed
aggregate-olm-edit ClusterRole Installed
olm Namespace Installed
global-operators operators OperatorGroup Installed
operators Namespace Installed
packageserver olm ClusterServiceVersion Installed
olm-operator-serviceaccount olm ServiceAccount Installed
catalogsources.operators.coreos.com CustomResourceDefinition Installed
system:controller:operator-lifecycle-manager ClusterRole Installed
As you can see from this output above, all of the components of the OLM are in the Installed status which means your Operator Lifecycle Manager is working properly.
Note: If you see an error, you can read the SDK guide to learn how to install OLM on your cluster.
(Optional) Troubleshooting OLM error
If you get an error similar to the following two, you need to uninstall a specific version
Error 1
$ operator-sdk olm status
FATA[0002] Failed to get OLM status: error getting installed OLM version (set --version to override the default version): no existing installation found
Error 2
$ operator-sdk olm status
FATA[0002] Failed to get OLM status: error getting installed OLM version (set --version to override the default version): no existing installation found
$ operator-sdk olm install
FATA[0005] Failed to install OLM version "latest": detected existing OLM resources: OLM must be completely uninstalled before installation
$ operator-sdk olm uninstall
FATA[0002] Failed to uninstall OLM: error getting installed OLM version (set --version to override the default version): no existing installation found
These errors indicate the need to uninstall a specific version. My OpenShift cluster is version 4.5.31_1531, so I had to uninstall version 0.16.1 using the following command.
$ operator-sdk olm uninstall --version 0.16.1
INFO[0009] Successfully uninstalled OLM version "0.16.1"
Next, I installed the same version, and checked the status:
$ operator-sdk olm install --version 0.16.1
INFO[0072] Successfully installed OLM version "0.16.1"
$ operator-sdk olm status
INFO[0004] Successfully got OLM status for version "0.16.1"
NAME NAMESPACE KIND STATUS
operators.operators.coreos.com CustomResourceDefinition Installed
operatorgroups.operators.coreos.com CustomResourceDefinition Installed
installplans.operators.coreos.com CustomResourceDefinition Installed
clusterserviceversions.operators.coreos.com CustomResourceDefinition Installed
olm-operator olm Deployment Installed
subscriptions.operators.coreos.com CustomResourceDefinition Installed
olm-operator-binding-olm ClusterRoleBinding Installed
operatorhubio-catalog olm CatalogSource Installed
olm-operators olm OperatorGroup Installed
aggregate-olm-view ClusterRole Installed
catalog-operator olm Deployment Installed
aggregate-olm-edit ClusterRole Installed
olm Namespace Installed
global-operators operators OperatorGroup Installed
operators Namespace Installed
packageserver olm ClusterServiceVersion Installed
olm-operator-serviceaccount olm ServiceAccount Installed
catalogsources.operators.coreos.com CustomResourceDefinition Installed
system:controller:operator-lifecycle-manager ClusterRole Installed
Once everything says installed, congratulations you are ready to start developing!
Conclusion
If you've installed these prerequisites, your environment is set up to develop an operator and deploy it to an OpenShift (or Kubernetes) cluster.