Deploying software with GitLab CI
I am working currently on setting up an automated deployment. Since my current project’s repo is located on GitLab, I decided to go with GitLab Continous Integration.
GitLab CI is integrated with GitLab and promises an easy-to-learn, multi-platform, multi-language way to test, build and deploy our code.
What we need
- Nothing more than a GitLab project
What we will do
- Set up continous integration with creating a config file called
- Configure GitLab project to use a Runner
In result every push to our git repository will trigger the Runner to automatically start the pipeline (the predefined jobs in our
.gitlab-ci.yml file), and which will output the result in our GitLab project’s pipelines page.
This tutorial shows briefly the procedure step by step: Quick start
Our first step is to write our
.gitlab-ci.yml file. This tutorial describes in detail how to do that: https://gitlab.com/help/ci/yaml/README.md#environment.
Noteworthy thing: Docker is integrated with GitLab CI, meaning that our Runner can use the Docker Engine to test and build our application by defining Docker Images. You can read more about Docker integration here: Docker integration
Runners run our yaml. We can create our own specific Runner, or we can used a Shared one. 2 Shared runners are available and set on GitLab projects by default. Read more about Runners: https://gitlab.com/help/ci/runners/README.md
Environments defined in our yaml will define where our code will get deployed.
If we want to deploy from our internal repo to a remote, external server we will need to use SSH keys Using SSH keys.
In addition I recommend this webcast to watch: Webcast Getting started with CI in GitLab (bad video quality).
Auto-deployment with GitLab CI checklist
- Create a Yaml file without any specific script, push it, check GitLab Piplines if it gets processed (Pending, Failed, Succesful) - by default GitLab will use its shared Runners
- Create your custom Runner if necessary, check if your Yaml gets processed with that Runner
- Set up an environment on GitLab Piplines > Environments, define it in Yaml
- Check if code gets deployed
- Create your Build and Test scripts, don’t forget to use the right Docker Images