Moodle on Azure

This repository pretend to act as an architectural guide to implement Moodle on Azure.

There are many ways to implement a cluster of Moodle, but we are going to cover two scenarios:

• Virtual Machine scale sets
• Azure Kubernetes Service

In both scenarios we are going to use a custom Moodle image based on Bitnami Moodle hardened resource: https://github.com/bitnami/bitnami-docker-moodle.

Versions used in this repository:

• MariaDB: 10.5.4
• MySQL: 5.7
• Moodle: 3.9.1 (3.9.1-debian-10-r17)

Configuring database

• Create the Azure Database for MySQL resource and then create a database and user in database to assign permissions.

  • Reference: https://docs.moodle.org/39/en/Installation_quick_guide

    Reference:

    CREATE DATABASE moodle DEFAULT CHARACTER SET utf8mb4 COLLATE utf8mb4_unicode_ci;
    GRANT ALL ON moodle.* TO 'moodleuser'@'%' IDENTIFIED BY 'Password.123';
    flush privileges;
    

Configuring Redis cache

• The custom Moodle image that is in this repo is prepared to use Redis cache, for this you need to create the resource Azure Cache for Redis and use the credentials inside of the file config.php.

• There are two places were you need to set configuration, one is in the code of config.php and second is in the plugin->cache configuration where you need to register the redis service.

Preparing Container Image

• It's possible to use the Dockerfile: https://github.com/robece/moodle-on-azure/blob/master/source/moodle/3.9.1-debian-10-r17/Dockerfile.

  • This file is configured to automatically install PHP-Redis, add custom files to the installation and grant permissions to use NFS mounted devices.

• There is a patch feature in the libmoodle.sh file to paste all the Moodle custom files: https://github.com/robece/moodle-on-azure/blob/master/source/moodle/3.9.1-debian-10-r17/rootfs/opt/bitnami/scripts/libmoodle.sh#L181.

  • This will help us to modify the resources any time the application is installed by the first time.

• After configure the container settings a new image will need to be push in the container registry (Dockerhub or Azure Container Registry) that will be used by the Virtual Machine or Kubernetes Service to download the image.

Note: If you don't know the right settings for the container try to run locally withe the Bitnami parameters.

Preparing Virtual Machine image for Virtual Machine scale set

• Create an Azure Virtual Network resource.

  • This virtual network will be use to connect all the resources like: VM, VMSS and Netpp.

• Create an Azure Virtual Machine resource based on Ubuntu Server 18.04 LTS.

  • This virtual machine will be used as the workspace where we are going to configure all the resources needed before create the final image used in the Virtual Machine scale set.

• Create an Azure NetApp Files resource.

  • This storage resource will be use to share the Moodle application files between machines.

• Create the Azure NetApp Files resource capacity pool and volume using the same Azure Virtual Network created before.

  • Azure NetApp Files volume will be used as a mounted device in the Azure Virtual Machine.

• Install NFS dependencies in the Ubuntu Azure Virtual Machine.

  • This dependencies can be found in the mounting instructions in the NetApp Files volume resource.

• Mount NFS device by modifying the /etc/fstab file in the Ubuntu Azure Virtual Machine.

  • These parameters required to can be found in the mounting instructions in the NetApp Files volume resource.

• Install Docker and Docker Compose.

  • Docker Compose will be use to run the Moodle container

• Create the docker-compose.yml file.

  • The docker-compose file must start like a service everytime a VM is started.
  • There is a sample reference here: https://github.com/robece/moodle-on-azure/blob/master/source/docker-compose/moodle/docker-compose.yml.

• Configure docker-compose service to start automatically.

  • There is a sample reference here:

• Once everything is working as expected, go to Azure Portal and shutdown the Virtual Machine, then create an image of the Virtual Machine.

  • This image will be use to create a new Virtual Machine scale set resource.

Preparing to deploy in AKS

• Create the Azure Kubernetes Service resource.

• Create an Azure NetApp Files resource.

  • This storage resource will be use to share the Moodle application files between machines.

• Create the Azure NetApp Files resource capacity pool and volume using the same Azure Kubernetes Service virtual network.

  • Azure NetApp Files volume will be used as a mounted device in the Azure Virtual Machine.

• Create the persistent volument and persistent volume claim resources.

  • There is a sample reference here: https://github.com/robece/moodle-on-azure/blob/master/source/aks/deployment-netapp-volume.yml.

• Create the horizontal pod autoscaling resource.

  • There is a sample reference here: https://github.com/robece/moodle-on-azure/blob/master/source/aks/deployment-hpa.yml.

• Deploy the Moodle custom image by using the Bitnami Helm chart.

  • There is a sample reference here:

    helm upgrade --install moodle bitnami/moodle --set image.repository=robecehub/moodle --set image.tag=1.0.38 --set replicaCount=1 --namespace=moodle --set externalDatabase.host=moodledb02.mysql.database.azure.com --set externalDatabase.port=3306 --set externalDatabase.database=moodle --set externalDatabase.user=moodleuser --set externalDatabase.password=Password.123 --set moodleSkipInstall=yes --set mariadb.enabled=false --set moodleUsername=admin --set moodlePassword=Password.123 --set [email protected] --set nodeSelector\.kubernetes\.io/os=linux --set service.type=LoadBalancer --set extraEnvVars[0].name=MOODLE_DATABASE_TYPE --set extraEnvVars[0].value=mysqli --set livenessProbe.enabled=false --set readinessProbe.enabled=false --set persistence.enabled=true --set persistence.existingClaim=pvc-nfs