Skip to content

Latest commit

 

History

History

MQTT

README.md

Lesson Edge - Deploy MQTT for the sensors to talk to

In this lab we are going to install and configure an MQTT Workload to run on k3s or RKE cluster

There are 2 methods provided to install MQTT, the Easy Method with Rancher and Helm, or the Manual Method which include the needed yaml files.

Prerequisites:

 Rancher
 Kubernetes Cluster
    Default StorageClass
    LoadBalancer

Easy Method via Rancher and Helm

1) Add k8s-at-home Repo

Select App & Marketplace -> Chart Repositories

2) Click Create to define a new Chart Repository

Name: k8s-at-home
Index url: https://k8s-at-home.com/charts/

You should now see the k8s-at-home Repo

3) Select Charts - You should now see Mosquitto as an available Chart

4) Install Mosquitto

Click on Values YAML and change the following items

service:
  main:
    type: LoadBalancer

5) Press Install and watch it deploy

Manual Method

1) Deploy Mosquitto using the local-path-provisioner

For x86

kubectl create -f mosquitto-deployment.yaml

For Raspberry Pi

kubectl create -f mosquitto-Raspberry-Pi-deployment.yaml

2) Deploy the Service

kubectl create -f mosquitto-service.yaml

3) Locate IP address of Mosquitto

kubectl get svc  mosquitto

Example:

NAME        TYPE           CLUSTER-IP      EXTERNAL-IP   PORT(S)          AGE
mosquitto   LoadBalancer   10.43.175.123   10.0.9.102    1883:30690/TCP   87s

This example tells us that Mosquitto is now available on 10.0.9.102 on port 1883