Skip to content

Latest commit

 

History

History
140 lines (101 loc) · 4.26 KB

README.md

File metadata and controls

140 lines (101 loc) · 4.26 KB
Raw

Kargo Advanced Example

This is a GitOps repository of an Kargo example that showcases advanced Kargo techniques and features. This example will create multiple Argo CD Applications and Kargo Stages with a pipeline to progress both Git and image changes through multiple Stages.

Features

  • A Warehouse which monitors both a container repository for new images and manifest changes in Git
  • Stage deploy pipeline with A/B testing
  • Promotion of Git changes and image tags
  • Verification with analysis of an HTTP REST endpoint
  • Argo CD Application syncing
  • Control flow Stage to coordinate promotion to multiple Stages
  • Rendered branches

Requirements

  • Kargo v0.9.x (for older Kargo versions, switch to the release-X.Y branch)
  • An Argo CD instance
  • GitHub and a container registry (GHCR.io)
  • git and docker installed

Instructions

  1. Fork this repo, then clone it locally (from your fork).

  2. Run the personalize.sh to customize the manifests to use your GitHub username and Argo CD destination:

    ./personalize.sh

  3. git commit the personalized changes:

    git commit -a -m "personalize manifests"
git push

  4. Create a guestbook container image repository in your GitHub account.

    The easiest way to create a new ghcr.io image repository, is by retagging and pushing an existing image with your GitHub username:

    docker buildx imagetools create \
  ghcr.io/akuity/guestbook:latest \
  -t ghcr.io/<yourgithubusername>/guestbook:v0.0.1

    You will now have a guestbook container image repository. e.g.: https://github.com/yourgithubusername/guestbook/pkgs/container/guestbook

  5. Change guestbook container image repository to public.

    In the GitHub UI, navigate to the "guestbook" container repository, Package settings, and change the visibility of the package to public. This will allow Kargo to monitor this repository for new images, without requiring you to configuring Kargo with container image repository credentials.

    change-package-visibility

  6. Download and install the latest CLI from Kargo Releases and Argo CD:

    ./download-cli.sh /usr/local/bin/kargo

  7. Login to Kargo and Argo CD:

    kargo login https://<kargo-url> --admin
argocd login <argocd-hostname>

  8. Create the Argo CD guestbook Project and Applications

    argocd proj create -f ./argocd/appproj.yaml
argocd appset create ./argocd/appset.yaml

  9. Create the Kargo resources

    kargo apply -f ./kargo

  10. Add git repository credentials to Kargo (replace <yourgithubusername> with your username).

    kargo create credentials github-creds \
  --project kargo-advanced \
  --git \
  --username <yourgithubusername> \
  --repo-url https://github.com/<yourgithubusername>/kargo-advanced.git

    As part of the promotion process, Kargo requires privileges to commit changes to your Git repository. Ensure that the given token has these privileges.

  11. Promote the image!

    You now have a Kargo Pipeline which promotes images from the guestbook container image repository, through a multi-stage deploy pipeline. Visit thekargo-advanced Project in the Kargo UI to see the deploy pipeline.

    pipeline

    To promote, click the target icon to the left of the dev Stage, select the detected Freight, and click Yes to promote. Once promoted, the freight will be qualified to be promoted to downstream Stages (staging, prod).

Simulating a release

To simulate a release, simply retag an image with a newer semantic version. e.g.:

docker buildx imagetools create \
  ghcr.io/akuity/guestbook:latest \
  -t ghcr.io/<yourgithubusername>/guestbook:v0.0.2

Then refresh the Warehouse in the UI to detect the new Freight.

Promoting Manifest Changes

To promote a manifest change, edit the contents under the base directory. For example, you modify guestbook-deploy.yaml with an additional environment variable:

        env:
        - name: FOO
          value: bar

Kargo will promote the environment variable in the same manner as with image tags.