Minikube - Kubernetes for developers

As described here, I decided to take a deeper look at minikube to run a Kubernetes cluster on my dev machine. These are the details and some of my learnings along the way.

Preparation

As usual, before starting any experiments, I want my MacBook up to date, so I make sure that I have the latest updates installed. At the time of writing, these are macOS Catalina version 10.15.5 and Xcode version 11.5. Then I make sure that brew and any installed packages are up to date as well:

brew update
brew upgrade
brew cask upgrade

In addition, make sure there are no issues. If there are, resolve them first:

brew doctor

Should give the following output when all is ready:

Your system is ready to brew.

Should you run into any issues, google is your friend…

If you are running windows, chocolatey is a decent alternative to brew.

If not already installed, make sure you have Docker Desktop up and running. This should also come with Hyperkit, a lightweight hypervisor for macOS. You can find other alternatives here.

Installation

Install minikube and kubectl:

brew install minikube
brew install kubectl

First Cluster

Create a cluster:

minikube start

Gives an output similar to:

😄  minikube v1.11.0 on Darwin 10.15.5
✨  Using the hyperkit driver based on user configuration
🆕  Kubernetes 1.18.3 is now available. If you would like to upgrade, specify: --kubernetes-version=v1.18.3
👍  Starting control plane node minikube in cluster minikube
🔥  Creating hyperkit VM (CPUs=2, Memory=2048MB, Disk=20000MB) ...
🐳  Preparing Kubernetes v1.14.7 on Docker 19.03.8 ...
🔎  Verifying Kubernetes components...
🌟  Enabled addons: default-storageclass, storage-provisioner
🏄  Done! kubectl is now configured to use "minikube"

❗  /usr/local/bin/kubectl is version 1.18.3, which may be incompatible with Kubernetes 1.14.7.
💡  You can also use 'minikube kubectl -- get pods' to invoke a matching version

This tells us that the default Kubernetes version that was used is 1.14.7, and that the cluster has 2 CPUs, 2 GB of memory and 20 GB of disk spaces at its disposable. It also tells us that the latest available version of Kubernetes is 1.18.3. Since this is my dev machine, I want to use the latest Kubernetes version, and I want to assign more resources. Remove the cluster:

minikube delete

Gives an output similar to:

🔥  Deleting "minikube" in hyperkit ...
💀  Removed all traces of the "minikube" cluster.

Create a new cluster as follows:

minikube start --kubernetes-version=v1.18.3 --cpus=4 --memory='8g' --disk-size='80000mb'

Gives an output similar to:

😄  minikube v1.11.0 on Darwin 10.15.5
✨  Using the hyperkit driver based on user configuration
👍  Starting control plane node minikube in cluster minikube
🔥  Creating hyperkit VM (CPUs=4, Memory=8192MB, Disk=80000MB) ...
🐳  Preparing Kubernetes v1.18.3 on Docker 19.03.8 ...
🔎  Verifying Kubernetes components...
🌟  Enabled addons: default-storageclass, storage-provisioner
🏄  Done! kubectl is now configured to use "minikube"

That is more like it. A local Kubernetes cluster up and running, and everything is up to date. If you want, you can open the kubernetes dashboard using:

minikube dashboard

The dashboard will be installed and opened in a browser tab.

Pause and resume

One very useful feature of minikube is that you can pause and resume your cluster. If you don’t need it for now, pause the cluster using:

minikube stop

Gives an output similar to:

✋  Stopping "minikube" in hyperkit ...
🛑  Node "minikube" stopped.

At a later point, you can resume it using:

minikube start --kubernetes-version=v1.18.3

Gives an output similar to:

😄  minikube v1.11.0 on Darwin 10.15.5
✨  Using the hyperkit driver based on existing profile
👍  Starting control plane node minikube in cluster minikube
🔄  Restarting existing hyperkit VM for "minikube" ...
🐳  Preparing Kubernetes v1.18.3 on Docker 19.03.8 ...
🔎  Verifying Kubernetes components...
🌟  Enabled addons: default-storageclass, storage-provisioner
🏄  Done! kubectl is now configured to use "minikube"

Notice that you do need to provide the Kubernetes version if you specified one different from the default. This command can also be used to upgrade the Kubernetes version if a newer one becomes available.

Multiple Clusters

Another very useful feature of minikube is the ability to run multiple clusters. You can run them at the same time to do multi-cluster experiments, or pause / resume them when switching between projects. Let’s first delete the previously created cluster:

minikube delete

Then create our first cluster:

minikube start -p cluster1 --kubernetes-version=v1.18.3

And then a second one:

minikube start -p cluster2 --kubernetes-version=v1.18.3

You can see the clusters using:

kubectl config get-contexts

Gives an output similar to:

CURRENT   NAME       CLUSTER    AUTHINFO   NAMESPACE
          cluster1   cluster1   cluster1
*         cluster2   cluster2   cluster2

The * marks the currently selected context. You can also see that using:

% kubectl config current-context
cluster2

You can switch between clusters by selecting their context:

% kubectl config use-context cluster1
Switched to context "cluster1".

Stop one of the clusters if you don’t need it for now:

% minikube stop -p cluster2
✋  Stopping "cluster2" in hyperkit ...
🛑  Node "cluster2" stopped.

Resume it when you need it again:

% minikube start -p cluster2 --kubernetes-version=v1.18.3
😄  [cluster2] minikube v1.11.0 on Darwin 10.15.5
✨  Using the hyperkit driver based on existing profile
👍  Starting control plane node cluster2 in cluster cluster2
🔄  Restarting existing hyperkit VM for "cluster2" ...
🐳  Preparing Kubernetes v1.18.3 on Docker 19.03.8 ...
🔎  Verifying Kubernetes components...
🌟  Enabled addons: default-storageclass, storage-provisioner
🏄  Done! kubectl is now configured to use "cluster2"

And delete it when you don’t need it anymore:

% minikube delete -p cluster2
🔥  Deleting "cluster2" in hyperkit ...
💀  Removed all traces of the "cluster2" cluster.

Next Steps…

Get more familiar with Kubernetes, follow some basic tutorials to deploy your first app here and find some more here. Again, google is your friend, there is plenty of content out there.

As for me, my next step will be to install Istio and make sure that my cluster can also be reached from other machines in my network. I’ll present the details in a following post…