cfl-summit-lab

Lab objectives:

In this lab session we will aim to complete three exercises:

1. Install Rancher on an RKE2 cluster
2. Provision a downstream RKE2 cluster using Rancher
3. Deploy an application using Fleet

Bonus rounds:

1. Take an etcd snapshot of the downstream cluster
2. Deploy an application or customize an existing addon using the Additional Manifests at the bottom of "Add-On Config" when editing the cluster
3. Restore the etcd snapshot taken above, choosing the third option to also revert the app deployment

Lab steps

Pre-work

• Obtain your node IPs and private SSH keys from the instructor, these will be your nodes to use throughout the lab session
  • There are two node IPs and private key files:
    • The node with r is the Rancher node used in exercise 1
    • The node with ds is the downstream node used in exercise 2
  • SSH username: ec2-user
  • sudo is available once logged in

1 - Install Rancher

Navigate to the Rancher install guide in the link below:

• https://ranchermanager.docs.rancher.com/getting-started/installation-and-upgrade/install-upgrade-on-a-kubernetes-cluster#install-the-rancher-helm-chart

1. SSH into the Rancher node (with r in the name)

2. Follow the steps to add the helm repo and create the cattle-system namespace. Choose the latest helm repository

3. From step 3 onwards, in this lab the "Rancher-generated TLS certificate" approach is recommended, where cert-manager will need to be installed (step 4)

4. For step 5, a hostname is needed for Rancher. You can use your own custom domain name, however for lab purposes, it is recommended to generate a quick hostname with .nip.io, just append this to your Rancher node Public IP: eg, 5.6.7.8.nip.io

   1. If you are unsure about the Public IP, use: curl ifconfig.io from the command line on the node
   2. Use this hostname for the --set hostname flag when running helm install, eg: --set hostname=5.6.7.8.nip.io
   3. Supply a bootstrap password to use when logging into Rancher for the first time with the --set bootstrapPassword flag
   4. Replace the <CHART_REPO> portion of the helm install command with latest, if latest was the repo chosen in the first step of this exercise
   5. Add the --set replicas 1 flag, to use just one Rancher pod, as we're running a single node for this lab

   Example command:

   helm install rancher rancher-latest/rancher --namespace cattle-system --set hostname=35.88.67.40.nip.io --set replicas=1
   

2 - Create a downstream cluster (Rancher Dashboard)

Navigate to the Downstream cluster configuration guide in the link below:

• https://ranchermanager.docs.rancher.com/reference-guides/cluster-configuration/rancher-server-configuration/use-existing-nodes

1. SSH into the downstream node (with ds in the name)
2. Follow the steps in the configuration guide (link above) from step 2 (Create the Custom Cluster)
3. At the end of step 2 you will use this node (with ds in the name) to run the provisioning command
4. Step 3 (Tag Resources) can be ignored

2a - Create a downstream cluster (terraform)

Pre-work:

• Clone this repo to the Rancher node (r): git clone https://github.com/rancherlabs/cfl-summit-lab.git
• Change directory to the 2a-downstream-cluster directory of the cloned repo, terraform is preinstalled on the node

1. Update the terraform.tfvars file to update the required variables (aws credentials, region, url, token, vpc details etc.)
2. Initialise the terraform modules: terraform init
3. Create the resources: terraform apply

3 - Deploy an application using Fleet

Instructions to deploy a webserver using fleet are provided in this link

• https://github.com/rancherlabs/cfl-summit-lab/blob/main/3-Deploy-an-application-using-Fleet/README.md

---

Bonus Rounds

1 - Take an etcd snapshot

Take an etcd snapshot of the downstream cluster(s). Details in the link below:

• https://ranchermanager.docs.rancher.com/how-to-guides/new-user-guides/backup-restore-and-disaster-recovery/back-up-rancher-launched-kubernetes-clusters#one-time-snapshots

2 - Deploy an application using a HelmChart (Add-On Config)

Navigate to Cluster Management in the Rancher dashboard

1. Click Edit Config on a downstream cluster
2. Navigate to Add-On Config under Cluster Configuration
3. Scroll down to Additional Manifest
4. Paste in a HelmChart, or Kubernetes manifest

Example: https://docs.rke2.io/helm#using-the-helm-crd

3 - Restore the etcd snapshot

Restore an etcd snapshot for the downstream cluster(s). Details in the link below:

• https://ranchermanager.docs.rancher.com/how-to-guides/new-user-guides/backup-restore-and-disaster-recovery/restore-rancher-launched-kubernetes-clusters-from-backup#restoring-a-cluster-from-a-snapshot

Choose the last option, to restore etcd, Kubernetes versions and cluster configuration

---

Cleanup (only if you completed exercise 2a)

cd 2a-downstream-cluster # if not in the directory
terraform destroy

---

Instructors only

See pre-work directory for preparation of the initial lab environment