Process notes for k3s online training

k3s-training.md

Prep

1. Generate a list of instances

• sbin/generate-instance-list.sh -n mak3r-rancher-training-k3s > instance-ids.json
• use a script to connect to those instances by list id
  • sbin/ssh-connect.sh -i 0 -k ~/.ssh/mak3r-private.pem -s

2. Get the cloud server public IP address and replace throughout the script
3. Prepare a local RPi server

---

Install and upgrade

Login

/Users/markabrams/dev/k3s-training/sbin/ssh-connect.sh -i 2 -k ~/.ssh/mak3r-private.pem -s

Demo install of k3s v0.3.0 (k8s version 1.13)

• curl -sfL https://get.k3s.io | INSTALL_K3S_VERSION=v0.3.0 sh -
• kubectl version
• k3s --version

Demo upgrade of k3s to latest (k8s version 1.14)

• curl -sfL https://get.k3s.io | INSTALL_K3S_VERSION=v0.7.0-rc4 sh -
• kubectl version
• curl -sfL https://get.k3s.io | INSTALL_K3S_VERSION=v0.7.0-rc4 INSTALL_K3S_EXEC="--write-kubeconfig-mode 644" sh -
• kubectl version
• k3s --version

Demo upgrade to latest k3s

• curl -sfL https://get.k3s.io | sh -
• kubectl version
• k3s --version

Demo uninstall

sudo /usr/local/bin/k3s-uninstall.sh

---

Multi-node clusters

Install server only

Login to the server (aws)

1. /Users/markabrams/dev/k3s-training/sbin/ssh-connect.sh -i 2 -k ~/.ssh/mak3r-private.pem -s
2. curl -sfL https://get.k3s.io | INSTALL_K3S_EXEC="--disable-agent --write-kubeconfig-mode 644 --tls-san 54.196.155.129" sh -
3. echo "export node_token=$(sudo cat /var/lib/rancher/k3s/server/node-token)"

• copy the output of the prior command

Setup the agents

• X86
  1. /Users/markabrams/dev/k3s-training/sbin/ssh-connect.sh -i 3 -k ~/.ssh/mak3r-private.pem -s
  2. curl -L -o k3s https://github.com/rancher/k3s/releases/download/v0.6.1/k3s
  3. chmod 755 k3s && sudo mv k3s /usr/local/bin/k3s
  4. Paste output of export node_token ... command
  5. sudo /usr/local/bin/k3s agent --server https://54.196.155.129:6443 --token "$node_token" >& k3s-agent.log &
• ARM
  1. /Users/markabrams/dev/k3s-training/sbin/ssh-connect.sh -i 0 -k ~/.ssh/mak3r-private.pem -s
  2. curl -L -o k3s https://github.com/rancher/k3s/releases/download/v0.6.1/k3s-arm64
  3. chmod 755 k3s && sudo mv k3s /usr/local/bin/k3s
  4. Paste output of export node_token ... command
  5. sudo /usr/local/bin/k3s agent --server https://54.196.155.129:6443 --token "$node_token" >& k3s-agent.log &

---

k3s on ARM

local node ARM

Login to the local node (RPi)

• ssh pi@k3s-calf.local

Install scripted

1. curl -sfL https://get.k3s.io | INSTALL_K3S_EXEC="--write-kubeconfig-mode 644" sh -
2. htop

Deployments

Create a deployment

1. vi nginx.yaml
2. kubectl apply -f nginx.yaml
3. kubectl get pods --all-namespaces
4. kubectl delete -f nginx.yaml

Demo manifests

1. vi nginx.yaml
2. sudo cp nginx.yaml /var/lib/rancher/k3s/server/manifests/.

---

k3d

• k3d create --name cluster01 --workers 3

1. KUBECONFIG="$(k3d get-kubeconfig --name='cluster01')"; kubectl get node
2. k3d delete -a

---

k3os

VBox gist

1. packer build k3os.json
2. vagrant up
3. vagrant ssh

---

Create clusters in Rancher

Use the UI to create the first cluster

• Retrieve the cluster kubectl command. Run it in the first cluster

Use the CLI to create additional clusters

1. rancher cluster create --import k3s01
2. rancher cluster import -q k3s01

• run output in the cluster to be imported

---

Cleanup

Cleanup of binary download types

1. sudo ps aux | grep k3s | awk '{print $2}' | sudo xargs kill
2. sudo rm /usr/local/bin/k3s

Cleanup of systemd scripted types

1. sudo /usr/local/bin/k3s-uninstall.sh