Below is a graph illustrating the IaC infrastructure.
graph LR
A[API developer Machine] --> B(Dockerfile)
B --> C(Container image)
C --> D(Azure Container Registry)
D --> E(Container App)
E --> F(Hosting app)
First of all I created API key from my account on https://openweathermap.org/ and created the link directly to my trigger:
I paste it in firefox and the program returned syntax in json:
Before starting the task related to Docker, I had to perform the following actions.
On my localhost I created x folder.
In~/x/weatherapp/frontend and ~/x/weatherapp/backend I used command:
npm i && npm start.
Then in ~/x/weatherapp/frontend and ~/x/weatherapp/backend I created Dockerfile's and do files as script to create docker.
Next in ~/x/weatherapp/ I created docker-compose.yml which can run a both of containers.
I include in docker-compose.yml required information about volumes I used:
npm run for frontend and backend are enabling hot reload.
To realize report the forecast e.g. a few hours from now I changed link to:
In fronted and backend API had much vulnerabilities and in consequences I used:
npm audit fix --force to up to date backend and frontend.
To work mocha properly, I needed to install additionally:
npm install -g mocha --save-dev in current folder frontend & backend
node node_modules/.bin/mocha --version mocha version
npm install axios required module for mocha
Then in folder testing I have created a openweathermap.test.js whch is checking valid city and return result of the test (screen in testing folder).
Execute command:
npx mocha openweathermap.test.js.
The First thing what I did was building three containers with equivalent command:
docker build -t weatherapp-backend:latest
docker build -t weatherapp-frontend:latest
docker build -t weather-data-provider:latest
When creating the infrastructure for a weather application, I decided to use a serverless solution.
To set up the weather service I have been choosen Azure portal. On my account I have been created resource group named x.
In this resource group i created Container registry openweathermap.
Locally, in te terminal, I executed several commands:
docker login operweathermap.azurecr.io
then execute tag & upload existing, local containers to the Azure Container registry:
docker tag weatherapp-backend openweathermap.azurecr.io/weatherapp-backend:latest
docker push openweathermap.azurecr.io/weatherapp-backend:latest
docker tag weatherapp-frontend openweathermap.azurecr.io/weatherapp-frontend:latest
docker push openweathermap.azurecr.io/weatherapp-frontend:latest
docker tag weather-data-provider openweathermap.azurecr.io/weather-data-provider:latest
docker push openweathermap.azurecr.io/weather-data-provider:latest
In Azure Container Apps service I created three container apps. I am hosting them on links below:
https://containerapp.purplegrass-e2305b4d.eastus.azurecontainerapps.io/ as weather-fronetend
https://wbackend.internal.purplegrass-e2305b4d.eastus.azurecontainerapps.io/ as weather-backend
https://wdataprovider.purplegrass-e2305b4d.eastus.azurecontainerapps.io/ as weather-data-provider
Before previous step I created Container Apps Environment service with name managedEnvironment-x-acf6 where I see that three container apps (screen on weatherapp folder).
I created ansible.yml file which automatically install on local machine the following services:
Docker & start docker service
weather application
mocha for testing.
I would like to propose an additional automation, whis is the automatic installation and starting of a container with Ubuntu 20.04 on the Proxmox virtualizer using:
ansible-playbook -i hosts ansible2.yml.
Then, by running necessary preliminary commands:
apt-get install ansible -y
apt-get install python3-pip -y
apt-get install python3-venv -y
python3 -m venv venv
pip install ansible -y
source venv/bin/activate
to start the ansible playbook:
ansible-playbook ansible1.yml -vvv
can be executed to install the remaining services.
Kod objęty licencją MIT – zobacz szczegóły