Kubernetes Cluster Setup Script

This script automates the setup of a Kubernetes cluster using Multipass for virtual machines and HAProxy for load balancing. It allows you to create a highly available Kubernetes cluster with customizable resources for master and worker nodes.

Prerequisites

• Multipass: Ensure Multipass is installed on your machine. See Multipass Install.
• kubectl: Install kubectl to interact with your Kubernetes cluster. See kubectl installation guide.
• jq: Install jq for JSON parsing. Run sudo apt-get install jq or follow the official guide.

References

• Multipass
• Docker Install
• Creating Highly Available Clusters with kubeadm
• Installing kubeadm
• CRI-O
• Weaveworks
• LUNXXtips YouTube Tutorials

Usage

1. Clone the Repository:

   git clone https://github.com/eliasmeireles/k8s.git
   cd k8s/learning/cluster-generate
   

2. Run the Script:

   ./setup
   

Options

Example

./scripts/setup --base-name dev-k8s --cluster-version v1 --num-masters 3 --num-workers 4 --master-mem 3G --worker-mem 6G

• This command will create a Kubernetes cluster named dev-k8s, with version v1, 3 master nodes, and 3 worker nodes. Master nodes will have 3GB of memory, and worker nodes will have 6GB of memory.

Network Setup

The script automatically applies a network plugin. For non-ARM architectures, Weave is used:

kubectl apply -f https://github.com/weaveworks/weave/releases/download/v2.8.1/weave-daemonset-k8s.yaml

For ARM architectures, Calico is applied:

kubectl apply -f https://docs.projectcalico.org/manifests/calico.yaml

Logging

• Logs for each execution are stored in the ./.temp directory with filenames based on the timestamp of execution.

Notes

• The script assumes a Unix-like environment.
• Ensure you have sufficient resources (CPU, memory, disk space) on your host machine to run the virtual machines.
• Each node created will have the necessary configuration and Kubernetes setup executed via a transferred script ( kube-config.sh).

Example Output

After completing the steps described in Master Node Setup, you will see the output below:

Installing kube-prometheus

To monitor your cluster, you can install kube-prometheus:

Make sure that you checked the expected cluster context

make apply-kube-prometheus

When all pods in the monitoring namespace are running, access the Prometheus dashboard:

Access the dashboard at http://monitor.v1.k8s.local and use the default credentials (admin/admin).

Cleanup

To delete all instances created by the script, use:

./scripts/delete-instances.sh <pattern>

Replace <pattern> with a keyword that matches the instance names to delete.

Author

Created by Elias Meireles.