Nov 25, 2023 3 min read MacOS

Running a local Kubernetes Cluster on an Apple Silicon MacBook

I wanted to test some stuff on my fairly new Mac, equipped with the M2 chip, and consequently, utilizes the arm64 architecture. I've bootstrapped a cluster using minikube at first, but due to some issues, switched to Docker desktop, which allowed me to "use Rosetta for x86/amd64 emulation on Apple Silicon" as I had issues with container images that were not built for this architecture. I could easily pull amd64 images using the platform flag, e.g.

docker pull IMAGE_NAME:TAG --platform linux/amd64

So I deployed stuff into my local cluster and ran into the following issue:

Events:
  Type     Reason     Age   From               Message
  ----     ------     ----  ----               -------
  Normal   Scheduled  18s   default-scheduler  Successfully assigned default/defectdojo-initializer-2023-11-25-23-16-64k2n to docker-desktop
  Normal   Pulling    18s   kubelet            Pulling image "defectdojo/defectdojo-django:latest"
  Warning  Failed     11s   kubelet            Failed to pull image "defectdojo/defectdojo-django:latest": no matching manifest for linux/arm64/v8 in the manifest list entries
  Warning  Failed     11s   kubelet            Error: ErrImagePull
  Normal   BackOff    11s   kubelet            Back-off pulling image "defectdojo/defectdojo-django:latest"
  Warning  Failed     11s   kubelet            Error: ImagePullBackOff

I tried to find an option to define the platform the container images should be for, without any success. But it looked like other people had similar issues

Change platform for kubernetes when pulling image

I am using an Apple M1, writing a Kubernetes YAML file to pull and deploy the image confluentinc/cp-zookeeper:6.2.4 the problem is, I cannot pull the image due to the difference in CPU architecture

Stack OverflowWibu

Minikube to deploy linux/amd64 images on M1 hardware

I know it’s possible to run locally on Apple’s M1 processor an amd64 image by using the --platform linux/amd64 flag on docker run. But how do I deploy that same image on a local Minikube cluster, r…

Stack Overflowalpha_hotel

In the end, I've stumbled upon Colima

Got right into setting it up and starting up

$ colima start --kubernetes --arch x86_64 --cpu 8 --memory 8 --runtime containerd

INFO[0000] starting colima
INFO[0000] runtime: containerd+k3s
INFO[0000] creating and starting ...                     context=vm
INFO[0242] provisioning ...                              context=containerd
INFO[0242] starting ...                                  context=containerd
INFO[0247] provisioning ...                              context=kubernetes
INFO[0248] downloading and installing ...                context=kubernetes
INFO[0273] loading oci images ...                        context=kubernetes
INFO[0309] starting ...                                  context=kubernetes
INFO[0337] updating config ...                           context=kubernetes
INFO[0338] Switched to context "colima".                 context=kubernetes
INFO[0339] done

Verify that everything is working as expected

$ kubectl get nodes

NAME     STATUS   ROLES                  AGE     VERSION
colima   Ready    control-plane,master   5m50s   v1.28.3+k3s2

I've tested my deployment again and the result was the following

Events:
  Type    Reason     Age    From               Message
  ----    ------     ----   ----               -------
  Normal  Scheduled  7m28s  default-scheduler  Successfully assigned default/defectdojo-initializer-2023-11-26-00-25-7ckz2 to colima
  Normal  Pulling    7m22s  kubelet            Pulling image "defectdojo/defectdojo-django:latest"
  Normal  Pulled     24s    kubelet            Successfully pulled image "defectdojo/defectdojo-django:latest" in 6m57.795s (6m57.795s including waiting)
  Normal  Created    24s    kubelet            Created container initializer
  Normal  Started    23s    kubelet            Started container initializer

The only kinda limiting factor right now is the performance...