Automating Image Deployment¶

OVERVIEW

This guide will walk you through different strategies for automating the deployment of your custom docker images to a ComputeStacks instance.

Using a Makefile¶

This makefile uses an environmental variable to store your ComputeStacks API credentials. If you will be running this from your local computer, I recommend using a tool such as direnv to manage your environmental variables.

ComputeStacks Authentication¶

Generate your API credentials from within the ComputeStacks interface and make them available to your local shell.

Using direnv

cd ~/my-project && echo 'export COMPUTESTACKS_AUTH_KEY="MY-CS-API-KEY:MY-CS-API-SEDCRET"' >> ~/.envrc
direnv allow .

Without any tool

export COMPUTESTACKS_AUTH_KEY="MY-CS-API-KEY:MY-CS-API-SEDCRET"

Create your Makefile¶

Replace https://my.computestacks.com/api/container_services/100/power/rebuild with the URL of your ComputeStacks installation, and 100 with the ID of your container service. You can easily find this by navigating to the service overview page (Projects → My Project → My Service) and looking in the URL.

.PHONY: build help run push

help: ## Help
  @awk 'BEGIN {FS = ":.*?## "} /^[a-zA-Z_-]+:.*?## / {printf "\033[36m%-30s\033[0m %s\n", $$1, $$2}' $(MAKEFILE_LIST)

.DEFAULT_GOAL := help

build: ## Build Image
  @docker build --progress plain -t my.registry.server/myimage:imagetag .

push: ## Deploy newly crated image
  @docker push my.registry.server/myimage:imagetag

deploy: ## Trigger rebuild of site
  @curl -X PUT -u "$(COMPUTESTACKS_AUTH_KEY)" -H "Content-Type: application/json" -H "Accept: application/json" https://my.computestacks.com/api/container_services/100/power/rebuild

Now, your process to build and deploy your image will be as simple as:

make build
make push
make deploy

Using Gitlab Actions¶

In this example, we will use Gitlab CI to build an deploy our docker containers.

Warning

Before proceeding: If you’re using the self-hosted version of Gitlab, be sure you already have Gitlab CI runners installed and working before proceeding. Learn More

Setup your Gitlab Project¶

If you have not done so already, create a new project in gitlab and clone the repo to your local computer. In that repo, setup your Dockerfile. Commit the file and push them to your project.

Navigate to Settings → CI/CD → Variables and create the following variables (enter the appropriate values based on your configuration):

Variable	Description
`CS_AUTH_KEY`	This is your CS `API KEY` and `API SECRET` in the format: `KEY:SECRET`
`CS_CONTROLLER`	The hostname of your CS controller. Example: `controller.example.net`
`CS_SERVICE_ID`	The ID of your container service.

Create your Gitlab CI file¶

In your local repo, create a file named .gitlab-ci.yml — in that file, place the following contents:

stages:
  - build
  - deploy

build-image:
  stage: build
  tags:
    - shell 
  variables:
    IMAGE_TAG: $CI_REGISTRY_IMAGE:$CI_COMMIT_REF_SLUG
  script:
    - docker login -u $CI_REGISTRY_USER -p $CI_REGISTRY_PASSWORD $CI_REGISTRY
    - docker build --progress plain -t $IMAGE_TAG .
    - docker push $IMAGE_TAG
    - docker logout $CI_REGISTRY
  only:
    - main

deploy-image:
  stage: deploy
  variables: {}
  tags:
    - shell
  script:
    - 'curl -X PUT -u "${CS_AUTH_KEY}" -H "Content-Type: application/json" -H "Accept: application/json" https://${CS_CONTROLLER}/api/container_services/${CS_SERVICE_ID}/power/rebuild'
  only:
    - main

Commit that file and push it to Gitlab.

Running your Gitlab CI Pipeline¶

You can now manually trigger your pipeline to run by navigating to CI/CI → Pipelines.

If you would like to automate when this ones, or build more advanced functionality, please see Gitlab’s documentation for more info.