Continuous deployment using GitHub actions

Simple pipeline for test, image, and deploy app onto Kubernetes cluster using GitHub action.

This pipeline is triggered by the creation of a release tag (e.g. v.1.2.3). It sets up the test environment and runs unit tests. If tests are successful, the pipeline creates configuration data, builds and publishes the image to the GitHub container registry, as well as deploys the resulting image using the simple deployment strategy and sends notifications on the progress of the entire pipeline to Discourt

This demo assumes yoo already have a Kubernetes cluster configured with Dapr. If not, consider the opinionated Dapr install in dapr-demos/setup.

Demo

To walk-through the demo steps, start by navigating to the the already deployed app and take a note of the release version and the deployment time (in UTC):

https://gitops.thingz.io/

Edit code

Next, edit the staticMessage variable in app/main.go to simulate developer making code changes:

Make sure to save your changes

const (
  greetingsMessage = "hello PDX"
)

Now, increment the version number variable (APP_VERSION) in the app/Makefile:

APP_VERSION ?=v0.1.5 # was v0.1.4

Sync changes

Add, commit, and push your local changes upstream:

make sync

This will git add, git commit, and git push your changes to GitHub

Create a release tag

When ready to make a release, tag it and push the tag to GitHub:

make tag

This will git tag it and git push origin your version tag to trigger to pipeline

Note, the GitHub pipeline takes about ~2 min from the time you tag it to when the new app is deployed. To monitor the results either check the GitHub notifications or watch the action execute the individual steps although that link will depend on your GitHUb username (e.g. mchmarny above).

View it

Once the pipeline is finished, you navigate again to the app.

https://gitops.thingz.io/

If everything went well, the new release version should reflect the change you made to the variable (APP_VERSION) in the app/Makefile and the deployment time (in UTC) should be also updated. And the Discord server should include the status notification.

If the changes are not there, check the GitHub Action to check on the status.

Setup Demo

Initial deployment

To setup the demo, first create the namespace:

kubectl apply -f config/space.yaml

If you have certs for the demo domain create a TLS secret:

kubectl create secret tls tls-secret -n gitops \
  --key certs/ingress-key.pem \
  --cert certs/ingress-cert.pem

Now apply the Dapr component and the rest of deployment:

kubectl apply -f component/
kubectl apply -f config/

When the command completed, check on the status:

kubectl get pods -n gitops

The response should include the gitops pod in status Running with container ready state 2/2:

NAME                      READY   STATUS    RESTARTS   AGE
gitops-5fb4d4d6f9-6m74l   2/2     Running   0          25s

One last check on ingress:

kubectl get ingress -n gitops

It should include gitops host as well as the cluster IP that's mapped in your DNS. If any of this sounds confusing, the the cluster setup instructions here.

NAME                   HOSTS              ADDRESS    PORTS   AGE
gitops-ingress-rules   gitops.thingz.io   x.x.x.x    80      19s

If everything went well, you should be able to navigate now to:

https://gitops.thingz.io

Kubernetes config

To enable GitHub action to deploy the built images to your cluster you'll first need to configure its context. If you already have authenticated to that cluster you can find this info in the .kube folder in your home directory. To ensure that the config has only information for that one cluster, it may be easier to simply export it from your managed Kubernetes provider.

Warning, the exported file has sensitive information, make sure to delete it after

For AKS for example:

az aks get-credentials --name demo2 --file sa.json

next, create GitHub secret (named KUBE_CONFIG) with the content of that file on the repo where the action will run.

That concludes the setup. You can navigate to the top of this readme and run the demo.

Disclaimer

This is my personal project and it does not represent my employer. While I do my best to ensure that everything works, I take no responsibility for issues caused by this code.

License

This software is released under the MIT

mchmarny/git-ops