DEV Community

LitmusChaos profile image Akash Shrivastava
Akash Shrivastava for LitmusChaos

Posted on

Setting up LitmusChaos on Raspberry Pi Cluster

#chaosengineering #raspberrypi #litmuschaos #devops

This blog is a guide on how to set up LitmusChaos on a Raspberry Pi cluster. This kind of setup can be used for development or testing purposes, as it is cheaper than cloud-based services, and it overcomes any limitations on your system.

LitmusChaos is a toolset to do cloud-native chaos engineering. It provides tools to orchestrate chaos on Kubernetes to help SREs find weaknesses in their deployments. SREs use Litmus to run chaos experiments initially in the staging environment and eventually in production to find bugs and vulnerabilities. Fixing the weaknesses leads to increased resilience of the system.

You can use this setup to see LitmusChaos in action, as well as this, can be used for development-level testing of services using Litmus.

Setting up a Raspberry Pi Cluster

This section is a guide on how to set up an RPi cluster to run Kubernetes with a Master and multiple Worker Nodes.

Hardware Required

For setting up the RPi Cluster, we need the following hardware (minimum requirement)

  1. Raspberry Pis (at least 2, the 4 GB variant will be good enough)
  2. Power Hub for powering the Raspberry Pis
  3. Ethernet Cable(s)
  4. Router (Optional Wi-Fi)
  5. 32 GB SD Card(s) (One for each RPi)
  6. MicroSD Card Reader (or MicroSD Slot on your Laptop)

Installing Operating System on SD Card

There are many Linux-based distros available for RPis, you can go with the RaspiOS Lite, the only drawback is that it is only available for 32bit systems. Considering that, you can choose Ubuntu 20.02 Server, which is also lightweight (not as much as RaspiOS) but it has been working fine. For this article, I will be using Ubuntu 20.02 Server.

Raspberry Pi provides an official image tool for installing the operating system on SD Card, but you can use any other tool as well. Download the Ubuntu image from here. Next, connect the SD card and open the image tool. Select the Choose OS option and then select the Custom Image option, select the ubuntu image you downloaded. Next, select the storage device and click on Write. This will take some time (from 5–20 minutes), and once done, continue the same process for all other SD Cards.

After this, insert the SD Cards into the Raspberry Pis and power them on.

Connecting RPis to Wifi (Optional)

If you want to use the RPis connected with Ethernet only then you can skip this step. Also if you have a mini-HDMI to HDMI converter, you don’t need an Ethernet cable to set up wifi, you can connect your RPis to a screen and follow the same process.

To connect your RPis to Wifi, you will have to first connect it with an Ethernet cable. Go to your router settings and get the IP address of the RPis. Then SSH into the RPis one by one and repeat the same step.

ssh ubuntu@<ip-addr-rpi>
Enter fullscreen mode Exit fullscreen mode

Note: The default password is ubuntu

You need to find the network interface name first

iw dev | grep Interface
Enter fullscreen mode Exit fullscreen mode

Now to connect to Wifi you have to edit the netplan configuration

sudo nano /etc/netplan/50-cloud-init.yaml
Enter fullscreen mode Exit fullscreen mode

Then add the following inside network block

wifis:
    <interface-name>:
        dhcp4: true
        optional: true
        access-points:
            "<your-wifi-ssid>”:
                password: "<your-wifi-password>"
Enter fullscreen mode Exit fullscreen mode

Save and exit the editor and then apply the new configuration

sudo netplan apply
Enter fullscreen mode Exit fullscreen mode

Now your device should be connected to wifi, you can check by

ip a
Enter fullscreen mode Exit fullscreen mode

Now, repeat the same process on all the Raspberry Pis and then you can disconnect the Ethernet cable.

Note: the IP address has changed.

Configuring Raspberry Pis for SSH

First, change the hostname of the Pis so they are easy to distinguish.

For master node

sudo hostname-ctl set-hostname kmaster
Enter fullscreen mode Exit fullscreen mode

For worker nodes

sudo hostname-ctl set-hostname knode<node number>
Enter fullscreen mode Exit fullscreen mode

Now, on your system create SSH keys and authorise them for the RPis by following these steps

Note: Following steps are to be followed on your system

  1. Create a .ssh directory if it doesn’t exist and cd into it 
mkdir .ssh && cd .ssh
Enter fullscreen mode Exit fullscreen mode

2. Use ssh-keygen to create SSH keys for master and all worker nodes, name the keys according to the hostname of the nodes so it’s easy to find.

3. Add the SSH keys to the ssh-agent

ssh-add kmaster
ssh-add knode<node number>
Enter fullscreen mode Exit fullscreen mode

4. Copy the ssh-keys to the RPis

_\# Master node_
ssh-copy-id -i ~/.ssh/kmaster.pub ubunut@<RPI-IP-ADDRESS>

_\# Worker node_
ssh-copy-id -i ~/.ssh/knode<number>.pub ubuntu@<RPI-IP-ADDRESS>
Enter fullscreen mode Exit fullscreen mode

If you had defined a static IP address for the RPis, you can use a hostname rather than an IP address

echo -e "<master node ip address>\\tkmaster" | sudo tee -a /etc/hosts
echo -e "<worker node 1 ip address>\\tknode1" | sudo tee -a /etc/hosts
Enter fullscreen mode Exit fullscreen mode

Now, try to login into the RPis to verify that everything is working fine

ssh ubuntu@kmaster
ssh ubuntu@knode1
Enter fullscreen mode Exit fullscreen mode

Installing Kubernetes on Raspberry Pi Cluster

=============================================

This section is a guide on how to install Kubernetes on Raspberry Pi Cluster with a Master and multiple Worker Nodes. We will be installing k3s because it is lightweight, you can install any other distribution as well.

Since we will be using Docker, follow the official docs to install, you can find them here.

Installing K3s Master

SSH into the master node

ssh ubuntu@kmaster
Enter fullscreen mode Exit fullscreen mode

Now install K3s

curl -sfL [https://get.k3s.io](https://get.k3s.io) | sh -s - --docker
Enter fullscreen mode Exit fullscreen mode

Verify that the installation was successful

sudo kubectl get nodes
Enter fullscreen mode Exit fullscreen mode

Note: You can check the k3s service to debug if the installation was not successful

sudo systemctl status k3s
Enter fullscreen mode Exit fullscreen mode

Installing K3s Nodes

On your system, run the following command to get the node token from the k3s master

MASTER\_TOKEN=$(ssh ubuntu@kmaster "sudo cat /var/lib/rancher/k3s/server/node-token")
Enter fullscreen mode Exit fullscreen mode

Now SSH into the node

ssh ubuntu@knode1
Enter fullscreen mode Exit fullscreen mode

Install K3s agent

curl -sfL [http://get.k3s.io](http://get.k3s.io) | K3S\_URL=https://kmaster:6443 K3S\_TOKEN=$MASTER\_TOKEN sh -s - --docker
Enter fullscreen mode Exit fullscreen mode

Verify that the K3s agent was installed successfully

sudo systemctl status k3s-agent
Enter fullscreen mode Exit fullscreen mode

Kubectlt

Install kubectl, a command-line interface tool that allows you to run commands against a remote Kubernetes cluster.

Now, create a config file to access the RPis K3s Cluster

mkdir -p $HOME/.kube/k3s
touch $HOME/.kube/k3s/config
chmod 600 $HOME/.kube/k3s/config
Enter fullscreen mode Exit fullscreen mode

Next, copy the k3s cluster configuration from the master node

ssh pi@kmaster "sudo cat /etc/rancher/k3s/k3s.yaml" **\>** $HOME/.kube/k3s/config
Enter fullscreen mode Exit fullscreen mode

Edit the k3s config file on the client machine and change the remote IP address of the k3s master from localhost/127.0.0.1 to kmaster

_\# Edit master config_
nano $HOME/.kube/k3s/config

_\# Search for the 'server' attribute located in -_
_\# clusters:_
_\# - cluster:_
_\#   server:_ [_https://127.0.0.1:6443_](https://127.0.0.1:6443) _or_ [_https://localhost:6443_](https://localhost:6443)
_#_
_\# Change 'server' value to_ [_https://kmaster:6443_](https://kmaster:6443)_
\# Do not change the port value_
Enter fullscreen mode Exit fullscreen mode

Now, export the _k3s_ config file path as the KUBECONFIG environment variable to use the config

export KUBECONFIG=$HOME/.kube/k3s/config
Enter fullscreen mode Exit fullscreen mode

Verify the setup

kubectl get nodes
Enter fullscreen mode Exit fullscreen mode

Installing LitmusChaos on Raspberry Pi Cluster

==============================================

This section is a guide on how to install LitmusChaos 2.0 on Raspberry Pi Cluster with K3s

For installation, we will be following their docs. There are two ways to install, one is by using helm, other is by applying the YAML spec file. We will be installing using the YAML spec file, you can follow the other one if you want by going through their docs.

kubectl apply -f [https://litmuschaos.github.io/litmus/2.0.0/litmus-2.0.0.yaml](https://litmuschaos.github.io/litmus/2.0.0/litmus-2.0.0.yaml)
Enter fullscreen mode Exit fullscreen mode

Note: You can find the latest version of litmus at docs.litmuschaos.io

Let’s verify that all the services are running and that there have been no issues

kubectl get pods -n litmus
kubectl get svc -n litmus
Enter fullscreen mode Exit fullscreen mode

You can now use the LitmusChaos dashboard by using this address

<master-node-ip>:<port>
Enter fullscreen mode Exit fullscreen mode

Change the with the master node IP and the with what is showing to you for the litmusportal-frontend-service external port value, the one after 9091: and then visit that address in your browser.

Add-Ons

=======

The /etc/hosts file sets to default after a restart, so you will have to keep adding the RPis IP every time you restart or you can run a startup script that will automatically set the values on every restart.

You can edit the bash profile file on your system to use this Kubeconfig and also add the ssh keys to the ssh-agent. In my system it was the /home/username/.profile file, it might differ in your system. I added these lines to the profile

eval $(ssh-agent)
ssh-add ~/.ssh/kmaster
ssh-add ~/.ssh/knode1export KUBECONFIG=$HOME/.kube/k3s/config
Enter fullscreen mode Exit fullscreen mode

Summary

=======

In this article, we first set up the Raspberry Pis cluster and then installed K3s on the cluster. After that, we installed LitmusChaos onto the K3s cluster. We can now proceed with injecting chaos using the portal. This kind of setup is beneficially for local development purposes, and you will be saving money on AWS servers.

You can join the LitmusChaos community on Github and Slack. The community is very active and tries to solve queries quickly.

I hope you enjoyed this journey and found the blog interesting. You can leave your queries or suggestions (appreciation as well) in the comments below.

Show your ❤️ with a ⭐ on our Github. To learn more about Litmus, check out the Litmus documentation. Thank you! 🙏

Thank you for reading

Akash Shrivastava

Software Engineer at Harness

Linkedin | Github | Instagram | Twitter

Top comments (0)

Read next

oussamalakhdar profile image

Kubernetes: An Overview of its Architecture and Main Components

Oussama Lakhdar -

sre_panchanan profile image

Mastering Traffic Management: A Comprehensive Istio Lab Guide

Panchanan Panigrahi -

markusmeyer13 profile image

How to create an Azure API Management Service with Developer Portal with CI/CD using Azure DevOps

Markus Meyer -

akbarnafisa profile image

Add Portainer Stack

Akbar Nafisa -