parcelLab SRE challenge solution

The repository contains solutions for all required tasks (minimum and optional ones).

1. Create a “Hello World” greetings service (REST-API) which responds with different salutations for customers (there won’t be further purpose for this API, just one GET endpoint is fine) Customer A wants to say “Hi”, customer B wants to say “Dear Sir or Madam”, customer C wants to say “Moin” and so on. Choose whatever programming language and framework you think might be the best fit Describe or implement the necessary things to ensure “good quality“ for your product
2. Automated docker build & push process with Github Actions plus deployment with helm to minikube.
3. BASH script for local build and push.

• Docker
• Helm

Docker build and run:

docker build -t parcel-lab .

docker run -p 5000:5000 parcel-lab

The parcel-lab python app will be available on http://localhost:5000

Build and push with the bash script. Example script run with arguments:

./build_push.sh --image-name my-app --image-tag v1.0 --registry https://index.docker.io/v1 --username my-docker-username --password my-docker-password --repository my-docker-repo

To repository has a pipeline for automated deployment of the helm chart to Minikube on every push in the main branch.

The pipeline requires two GitHub secrets to be set for the repository. It can be done on a repository or organization level.

• DOCKERHUB_USER (your docker registry user)
• DOCKERHUB_PASS (your docker registry password)

https://github.com/mmpetarpeshev/parcelLab-SRE/actions

The repository contains the helm chart for the parcelLab python app. The helm chart supports the creation of the following resources:

* deployment  
* hpa  
* ingress  
* pvc  
* service  
* service account  

The chart could be installed with the following command:

helm instal --values charts/parcel-lab/values.yaml parcel-lab ./charts/parcel-lab --dependency-update

Curl commands to test all customer's endpoints:

curl http://localhost:5000/greet/customerA

curl http://localhost:5000/greet/customerB

curl http://localhost:5000/greet/customerC

Simulate error with random wrong customer. For example:

curl http://localhost:5000/greet/customerD

Every pipeline run performs a test curl request for each customer endpoint and prints the result.

It can be seen under the Test Customer A/B/C Endpoint steps.

1. Pipeline Enhancement: Consider splitting the test steps into a separate pipeline as they expand.
2. Traefik Ingress: Explore exposing the service with Traefik Ingress (template available in the chart). Note: Traefik deployment to Minikube requires additional work.
3. Resource Creation Enhancement: Improve conditional creation of resources for better resource management.