Skip to content

Latest commit

 

History

History
91 lines (63 loc) · 4.48 KB

README.md

File metadata and controls

91 lines (63 loc) · 4.48 KB
Raw

Instructions for using this Platform NodeJS Repo Template for a NodeJS exposing an endpoint (e.g. API)

Use this template when creating a new NodeJS project that exposes an HTTP endpoint.

It provides basic files and configurations that will allow you to get up to speed when building a platform-deployable NodeJS application.

👀 This is not exactly what you need?

In case you need a different kind of application (e.g. static hosting, Cron-triggered application, etc.) and you can't find a template for it, please consider forking this repo.

Setup Deployment Pipeline

Step 1: Configure Project Files

Configure the following files.

Reference: TDA Dockerfile best practices.

TL;DR

In all files provided by this template, replace all of the following occurences as listed:

  • <APP_NAME> => your app's repo name
  • <NAMESPACE> => with a namespace for the Kubernetes location (so that it can be grouped with related applications)
  • <GROUP_ID> => by the output of python -c 'import random; print(random.randint(10000, 60000))' (run it in the terminal)

Configure Dockerfile

  1. Replace <GROUP_ID>:
    1. Generate a random group id by running python -c 'import random; print(random.randint(10000, 60000))' in the terminal
    2. Copy the generated id and replace <GROUP_ID> with it.
  2. Replace <APP_NAME> with your app name

Configure tdaci.env

In tdaci.env:

export TDACI_NAMESPACE=default				# Enviroment used, default will use python3
export MY_SERVICE_VERSION=0.0.1 			# Version of the application
export MY_RELEASE_BRANCH=release			# (Leave as is)
export MY_SERVICE_NAME=gb-services/<APP_NAME>		# Service name
export MY_K8S_NAMESPACE=<NAMESPACE>			# Kubernetes namespace where the service is located
export MY_K8S_DIR=<APP_NAME>				# Directory where the service is located inside namespace

Replace every occurence of <APP_NAME> and <NAMESPACE> with the app name and the namespace you want it to be listed in, respectively.

Step 2: Request a "CICD New Pipeline" @ #gb_ops_zähler on Slack

  1. Go to the #gb_ops_zähler channel on Slack

  2. Use the channel workflow shortcut to make a request for a new "CICD New Pipeline" workflow- shortcut

    1. Provide link to the repo
    2. Choose "GB" as the K8S Cluster group
    3. Add to: "GOCD (Build) and ARGOCD (Deploy)" for the full build and deployment pipeline
    4. Under Additional information add details about the application's context, technical and access requirements, interactions with API, etc. will help the @ops team follow-up on specifics and eventual manual configurations.

    example-form

  3. Submit request

  4. Prepare the secrets (e.g. AWS user credentials). You'll have to share them with the prompting member of the @ops team. Important: remember to delete secret-related messages on slack once they have been received by the other party.

Step 3: Test the pipeline

When set up properly, the pipeline should be triggered when pushing commits to the develop branch. (Production branch solution name not known at the moment of writing this.)

Once the deployment pipeline works properly, you may delete this whole section (Setup Deployment Pipeline).

Run and develop locally

npm run dev

Test in Docker

Test your application works within Docker before pushing your commits to the CI/CD pipeline:

  1. Make sure you have docker installed and running on your machine: https://docs.docker.com/get-docker/
  2. Add .env file with following content: 
    PORT=3000
    Important note: make sure environment variable files are always ignored by git.
  3. Run docker using the docker-compose.yml configuration: 
    docker compose up
  4. The endpoint should now be available at https://localhost:3000/

Important note on updating/changing source code locally: the following sequence of commands will build a docker image and run it. Make sure you understand, that changes in the code have to be rebuilt and run again. Existing images will have to be teared down (docker compose down) to make place for your changes. We reccomend using the Docker desktop dashboard application to get an overview of instances available/running on your machine.