minikube
minikube copied to clipboard
kubernetes
Katakoda is using a very old minikube v1.18
What Happened?
the katakoda on Kubernetes website is using a very old minikube and it needs to be updated. if anyone can give a hand in finding out how to update it, this task is available to be picked up
https://kubernetes.io/docs/tutorials/hello-minikube/
I presume any k8s upgrade would also include a minikube upgrade.
They are released independently, kubernetes doesn't even test with minikube (or with docker) anymore.
What I meant is, since most k8s live tutorials use minikube as well, it'd make sense to upgrade them together in the katacoda image. (Assuming there's anyone at katacoda/oreilly who could still do the upgrade)
Sure, it is the same task.
I still think it would be a good idea to have solution that could run both in a cloud provider as well as in your own laptop. There is also room for alternative scenarios such as multi-node, as a step two. Especially now when they can share one VM (running two node containers)
Except for some quirks, this is an approximation of an updated web UI:
As compared with the one that is currently in the release notes/tutorials:
As mentioned before, it would be nice to get rid of all the misleading spam from the output:
❗ Using the 'containerd' runtime with the 'none' driver is an untested configuration!
❗ With --network-plugin=cni, you will need to provide your own CNI. See --cni flag as a user-friendly alternative
config.WantNoneDriverWarning
❗ The 'none' driver is designed for experts who need to integrate with an existing VM
💡 Most users should use the newer 'docker' driver instead, which does not require root!
📘 For more information, see: https://minikube.sigs.k8s.io/docs/reference/drivers/none/
redundant commands (HOME=/home/anders.linux)
❗ kubectl and minikube configuration will be stored in /home/anders.linux
❗ To use kubectl or minikube commands as your own user, you may need to relocate them. For example, to overwrite your own settings, run:
▪ sudo mv /home/anders.linux/.kube /home/anders.linux/.minikube $HOME
▪ sudo chown -R $USER $HOME/.kube $HOME/.minikube
💡 This can also be done automatically by setting the env var CHANGE_MINIKUBE_NONE_USER=true
One improvement over the previous "none" version is that it no longer needs to run as root.
The new version of minikube uses sudo where needed, instead of running everything as root.
Running with the "docker" driver is a more advanced scenario, but good for multi-node ?
Assuming that it is OK to overprovision, otherwise one might as well start up two node VMs.
Here are the issues that you run into, when you try to run the "none" driver (in lima)
- minikube_1.28.0-0_amd64.deb
❌ Exiting due to GUEST_MISSING_CONNTRACK: Sorry, Kubernetes 1.25.3 requires conntrack to be installed in root's path
Fix: sudo apt install conntrack
🤦 StartHost failed, but will try again: creating host: create: precreate: exec: "docker": executable file not found in $PATH
Workaround: sudo ln -s nerdctl /usr/local/bin/docker
❌ Exiting due to RUNTIME_ENABLE: update sandbox_image: /bin/bash -c "sudo sed -e 's|^.*sandbox_image = .*$|sandbox_image = "registry.k8s.io/pause:3.8"|' -i /etc/containerd/config.toml": exit status 2
stdout:
stderr:
sed: can't read /etc/containerd/config.toml: No such file or directory
Fix: sudo mkdir -p /etc/containerd && sudo containerd config dump | sudo tee /etc/containerd/config.toml
❌ Exiting due to RUNTIME_ENABLE: Temporary Error: sudo crictl version: exit status 1
stdout:
stderr:
sudo: crictl: command not found
Workaround: https://github.com/kubernetes-sigs/cri-tools/releases/tag/v1.25.0
- crictl-v1.25.0-linux-amd64.tar.gz
💢 initialization failed, will try again: apply cni: cni apply: copy: chmod /etc/cni/net.d/1-k8s.conflist: permission denied
Fix: sudo chmod 755 /etc/cni /etc/cni/net.d
"RunPodSandbox from runtime service failed" err="rpc error: code = Unknown desc = failed to setup network for sandbox \"fec81ef612493011775ba2ec285fc375f79fa2e6671e8d6f2b6fcb47c5652130\": plugin type=\"loopback\" failed (add): failed to find plugin \"loopback\" in path [/opt/cni/bin]"
(there are some plugins for nerdctl, in /usr/local/libexec/cni)
Workaround: https://github.com/containernetworking/plugins/releases/tag/v1.1.1
- cni-plugins-linux-amd64-v1.1.1.tgz
Happy face, with all issues above fixed:
$ limactl start
😄 minikube v1.28.0 on Ubuntu 22.04 (kvm/amd64)
✨ Using the none driver based on user configuration
👍 Starting control plane node minikube in cluster minikube
🤹 Running on localhost (CPUs=4, Memory=3919MB, Disk=99053MB) ...
ℹ️ OS release is Ubuntu 22.04.1 LTS
📦 Preparing Kubernetes v1.25.3 on containerd 1.6.8 ...
▪ kubelet.resolv-conf=/run/systemd/resolve/resolv.conf
▪ Generating certificates and keys ...
▪ Booting up control plane ...
▪ Configuring RBAC rules ...
🔗 Configuring bridge CNI (Container Networking Interface) ...
🤹 Configuring local host environment ...
🔎 Verifying Kubernetes components...
▪ Using image gcr.io/k8s-minikube/storage-provisioner:v5
🌟 Enabled addons: default-storageclass, storage-provisioner
💡 kubectl not found. If you need it, try: 'minikube kubectl -- get pods -A'
🏄 Done! kubectl is now configured to use "minikube" cluster and "default" namespace by default
Still need --preload=false, if not using ISO.
This would also be a nice improvement:
sudo ln -s minikube /usr/bin/kubectl
Using the docker runtime would look similar.
$ limactl start --name default template://docker-rootful
😄 minikube v1.28.0 on Ubuntu 22.04 (kvm/amd64)
✨ Using the none driver based on user configuration
👍 Starting control plane node minikube in cluster minikube
🤹 Running on localhost (CPUs=4, Memory=3919MB, Disk=99053MB) ...
ℹ️ OS release is Ubuntu 22.04.1 LTS
🐳 Preparing Kubernetes v1.25.3 on Docker 20.10.21 ...
▪ kubelet.resolv-conf=/run/systemd/resolve/resolv.conf
▪ Generating certificates and keys ...
▪ Booting up control plane ...
▪ Configuring RBAC rules ...
🔗 Configuring bridge CNI (Container Networking Interface) ...
🤹 Configuring local host environment ...
🔎 Verifying Kubernetes components...
▪ Using image gcr.io/k8s-minikube/storage-provisioner:v5
🌟 Enabled addons: default-storageclass, storage-provisioner
💡 kubectl not found. If you need it, try: 'minikube kubectl -- get pods -A'
🏄 Done! kubectl is now configured to use "minikube" cluster and "default" namespace by default
🐳 Exiting due to NOT_FOUND_CRI_DOCKERD:
Workaround: https://github.com/Mirantis/cri-dockerd/releases/tag/v0.2.6
- cri-dockerd_0.2.6.3-0.ubuntu-jammy_amd64.deb
I think that SIG Docs is still trying to find alternative hosting, before Katacoda shuts down (tomorrow)
- https://github.com/kubernetes/website/issues/33936#issuecomment-1357005866
It is not clear if there will be a new deployment of the old version (1.18), or of the new version (1.28) ?
Client Version: version.Info{Major:"1", Minor:"20", GitVersion:"v1.20.2", GitCommit:"faecb196815e248d3ecfb03c680a4507229c2a56", GitTreeState:"clean", BuildDate:"2021-01-13T13:28:09Z", GoVersion:"go1.15.5", Compiler:"gc", Platform:"linux/amd64"}
Server Version: version.Info{Major:"1", Minor:"20", GitVersion:"v1.20.2", GitCommit:"faecb196815e248d3ecfb03c680a4507229c2a56", GitTreeState:"clean", BuildDate:"2021-01-13T13:20:00Z", GoVersion:"go1.15.5", Compiler:"gc", Platform:"linux/amd64"}
There is support for Kubernetes in Killercoda, but I don't think that it is using minikube ? (just kubeadm)
It does upgrade the OS version from Ubuntu 18.04 LTS to Ubuntu 20.04 LTS though, always something.
Client Version: version.Info{Major:"1", Minor:"25", GitVersion:"v1.25.3", GitCommit:"434bfd82814af038ad94d62ebe59b133fcb50506", GitTreeState:"clean", BuildDate:"2022-10-12T10:57:26Z", GoVersion:"go1.19.2", Compiler:"gc", Platform:"linux/amd64"}
Server Version: version.Info{Major:"1", Minor:"25", GitVersion:"v1.25.3", GitCommit:"434bfd82814af038ad94d62ebe59b133fcb50506", GitTreeState:"clean", BuildDate:"2022-10-12T10:49:09Z", GoVersion:"go1.19.2", Compiler:"gc", Platform:"linux/amd64"}
Made separate issues for the output spam:
- #15576
- #15577
See https://github.com/kubernetes/minikube/issues/15097#issuecomment-1345982947
It will still complain about "none", and about containerd.
But those can be worked around or fixed, and are not bugs.
❗ The 'none' driver is designed for experts who need to integrate with an existing VM
❗ Using the 'containerd' runtime with the 'none' driver is an untested configuration!
Especially ironic since containerd was the only runtime for 1.26.0
The support for Docker and for CRI-O was a bit late, and didn't arrive until some weeks later.
But it is true that it still buggy in minikube, due to it requiring a docker binary since forever.
- #10908
- #5549
https://kubernetes.io/blog/2023/02/14/kubernetes-katacoda-tutorials-stop-from-2023-03-31/