Application Setup and Deployment¶

1. Create a new Gitlab Project¶

Create the project¶

This is simply done by heading to the right group and clicking on create new project Architecture Diagram

Choose project format¶

Here it's recommended to use a Create blank project Architecture Diagram

Fill the project informations¶

Here you just need to enter your Project name No need to change the other settings Architecture Diagram

Project creation successful¶

Here the project has been created successfully with the name test-project

2. Add required merge request template¶

In order to have a default template to be used automatically in merge requests body

We reccommend that you add this default template. To do that simply

Create this folder path `.gitlab/merge_request_templates
Create a file named Default.md

Copy/Paste the following content:

# Fixes #<ticket_number>
<!-- Replace <ticket_number> with the relevant issue or ticket number -->

# Description
<!-- Briefly describe what this MR changes and why -->

# How to Test
<!-- Step-by-step instructions for testing this MR -->
1. Step one
2. Step two
3. Step three

3. CI/CD Pipeline Setup¶

Prerequisites:¶

You need to have a Dockerfile capable of packaging your application.
The Dockerfile needs to be built and tested locally to avoid any confusion with the pipeline later on.

Setup¶

In order to have an automated Gitlab CI/CD you have to set it up by creating a file named .gitlab-ci.yml at the root directory of your project.

Simple CI/CD Setup (one environment)¶

If you want to deploy a simple CI/CD setup that deploys to a single environment you need to:

Copy/Paste this configuration into the previously created .gitlab-ci.yml

stages:
  - build
  - trivy_scan
  - deploy

include:
  - component: $CI_SERVER_FQDN/components/docker/build@1.0.25
    inputs:
      stage: build
      branches: "main"
      tag: imagesbuilder 
  - component: $CI_SERVER_FQDN/components/docker/scan@1.0.25
    inputs:
      stage: trivy_scan
      branches: "main"
      tag: imagesbuilder
  - component: $CI_SERVER_FQDN/components/docker/deploy@1.0.25
    inputs:
      branches: "main" # is the branch that triggers the deployment
      environment: "production" # is the environment name 
      stage: deploy # deploy stage
      tag: <target_runner> # Runner in which VM you want to deploy the environment

- This pipeline have 3 stages mentioned in the beggning of the file which are build, scan and deploy - Do no change the build and scan steps - In the 3rd step comes the deployment, you need to adjust the tag <target_runner> that will specify in which VM your application will be deployed - More informations about each step can be found here - If you have another environment let's say staging you can add another step in the end of the file like this:

 - component: $CI_SERVER_FQDN/components/docker/deploy@1.0.25
   inputs:
     branches: "staging" # is the branch that triggers the deployment
     environment: "staging" # is the environment name 
     stage: deploy # deploy stage
     tag: <target_runner> # Runner in which VM you want to deploy the environment

2. Make sure to create a docker-compose-<environment_name>.yml in the root directory 3. Copy/Paste the following configuration and replace <application_name> with your application name:

version: '3.8'

services:
  <application_name>_<environment_name>:
    image: ${CI_REGISTRY}/${CI_IMAGE_NAME}:${VERSION}
    container_name: <application_name>_<environment_name>
    restart: always
    env_file: 
      - .env.<environment_name>
    ports:
      - "<gateway_exposed_port>:<your_docker_exposed_port>"
    networks:
      - <application_name>_<environment_name>
networks:
  <application_name>_<environment_name>:
    driver: bridge

NOTE: - if you have another environment you need to create another docker-compose-<environment_name>.yml matching the environment. let's say stagingenvironment you will have a docker-compose-staging.yml - Make sure to adjust the ports: - <gateway_exposed_port>: this can be chosen to anything you want, just avoid any ports already used in this file - <your_docker_exposed_port>: Your docker exposed port, this is usually the port in which your application runs

Define your environment/environments in Gitlab
You have to go to in the sidebar to Operate > Environments
Click into New environment and define the required informations
Do the same if you have other environments
CI/CD Variables Now you should be able to define your application variables depending on each environment.
For that you need to go in the sidebar to Settings > CI/CD
In the Project variables sections click on Add variable
Leave the enviroments to All and in the key enter CI_IMAGE_NAME in the value enter your wished image name
Now you can define your other environment variables which are used by your application.
- Make sure they all start with CI_ in the key
- Should the right environment if the variable is depended by the enviroment if not you can leave it to All which will be used with all environments
- Make sure to click at Masked and hidden if your variable holds any secrets/passwords
- save your variable, and repeat the process to all your other variables
WARNING:
The <environment_name> mentioned in docker-compose file and in the gitlab-ci file needs to be exact same.

4. Domain exposition¶

Make sure to register your application domain to point to this IP 198.50.212.113
Go to the gateway.yml file

Add your entry by Copy/Paste this configuration:

  - domain: <your_registred_domain>
    target: <internal_vm_ip_address>:<gateway_exposed_port>

internal_vm_ip_address: this is the VMs in which your application has been deployed to (dependent on the <runner_tag> you mentioned in the gitlab-ci configuration)
gateway_exposed_port: this is the same port you specified in you docker-compose-<environment_name>.yml file
Do the same thing if you have other environments
Save and push the file, after a while (CI/CD of the gateway finishes), you will be able to access your application with your <your_registred_domain>