README.md

xOpera deploy of simple thumbnail generator to Azure with Ansible

Table of Contents

• Purpose
• Functionality
• Quick test and deploy
• Deployment instructions
  • Prerequisites
  • Roles and deployment
    • Container deploy
    • Function deploy
    • Container notification
  • Setting inputs
  • Running with xOpera

Purpose

The repository implements image resize functionality deployment with xOpera for Azure Platform.

Functionality

The main functionality of image-resize is to create thumbnails from the source image. Source image must be uploaded into source bucket and then three thumbnails will be created and saved to another bucket.

Quick test and deploy

If you want to test this solution immediately run the following commands:

📝 I you face any problems consider reading the whole manual!

# install Azure CLI and try to login in your account
curl -sL https://aka.ms/InstallAzureCLIDeb | sudo bash
az login

# install prerequisites
cd azure
python3 -m venv .venv
source .venv/bin/activate
pip install --upgrade pip
pip install ansible
pip install opera

# setup azure function tools
wget -q https://packages.microsoft.com/config/ubuntu/19.04/packages-microsoft-prod.deb
sudo dpkg -i packages-microsoft-prod.deb
sudo apt-get update
sudo apt-get install azure-functions-core-tools
rm packages-microsoft-prod.deb

# run xOpera service (make sure to setup the env vars in yml file)
opera deploy -i inputs.yml resize_service_opera_v1_3.yml

Deployment instructions

The next part explains the deployment of this solution in detail.

Prerequisites

This topic explains prerequisites to run xOpera deploy via Azure. If you want to test it immediately go to quick test

To be able to run these roles you should use virtual environment, install the latest Ansible version.

python3 -m venv .venv
. .venv/bin/activate
pip install --upgrade pip
pip install ansible

Next, you can install opera orchestrator to your virtual environment from here. Basically you need to install it using pip install opera.

You should also install and sign in with Azure CLI using az login command. This is important because user gets authenticated via cli right now.

The next thing is to set up the Azure function tools using the commands below:

wget -q https://packages.microsoft.com/config/ubuntu/19.04/packages-microsoft-prod.deb
sudo dpkg -i packages-microsoft-prod.deb
sudo apt-get update
sudo apt-get install azure-functions-core-tools
rm packages-microsoft-prod.deb

Roles and deployment

There are 3 playbooks separated into folders. Usage is explained in the table below:

Role	Purpose
container-create	Creates storage account and necessary container on specified Azure storage account
function-deploy	Creates FunctionApp and deploys a new function to Azure portal using Azure CLI
container-notification	Creates an event subscription trigger for function

Container deploy

This part is used to create two containers (buckets) - original and resized on Azure Storage Account. Original container is used to upload images and then resized images get stored into resized container. This part also includes creating a whole new storage account.

Function deploy

Azure functions first need Azure Function App. Here we create a new Function App and then we deploy our Azure function by calling funct functionapp publish which deploys all functions stored in functionApp folder (these are visible in playbooks/function-deploy/functions). Currently there is a function called image-resize-function which is used for creating thumbnails.

Container notification

This part sets container notification for incoming container so that when a new image is uploaded that would trigger specified function from the functionApp. So here we use Azure EventGrid by creating event subscription.

Setting inputs

You can modify values of variables stored in inputs.yml to set the suitable parameters like IPs, auth params, container names, paths etc. Currently the following values can be modified:

Inout (variable)	Purpose	Example
host_ip	IP address of your target machine for execution	localhost
location	Location for the created resources	westeurope
resource_group_name	Name of the resource group	Resource-Group
storage_account_name	The name of the new storage account	mystorageaccount
container_in_name	The name of the incoming Azure blob storage container	original
container_out_name	The name of the container with results	resized
function_app_name	Name of the new Function App	MyFunctionApp
function_name	Name of the new Azure function - should be the same as folder name of the function	image-resize-function
event_subscription_name	Name of event subscription trigger that will be created	thumbnail-function-trigger

The file with inputs is later passed to xOpera at the beginning of the orchestration.

Running with xOpera

You can run xOpera resize service with opera deploy -i inputs.yml resize_service_opera_v1_3.yml