This repository is part of an article series

Introduction

This repository contains two applications:

• A Webhook payload delivery service for local development
• A Telegram bot that logs GitHub Actions workflows

I will write the second part of the article once I have more time, but feel free to read the first one.

bot-demo.mov

Setup

• Add a .env by cp .env.sample .env

# this can be found when creating your bot through BotFather
# https://core.telegram.org/bots/tutorial#obtain-your-bot-token
TELEGRAM_BOT_TOKEN=
# this can either be found through the browser URL, it should have a format liks "-1234567890"
# or you can use https://telegram.me/rawdatabot
TELEGRAM_CHAT_ID=

# this is for dev environment + it is explained on the article how it works
DEV_CLI_SOURCE_URL=https://your-service.com # e.g. https://smee.io/
DEV_CLI_TARGET_URL=http://web:3000/webhook

# you must have a github app created with the Webhook URL pointing to your server
# this will be the URL of the deployed webserver
# https://your-server.com/webhooks - for prod
# https://your-server.com/channel/<any-wildcard> - for dev
# more info on how to generate the key below
GITHUB_APP_PRIVATE_KEY=

# the APP_ID and APP_INSTALLATION_ID can be found directly in the Github App page
GITHUB_APP_ID=
GITHUB_APP_INSTALLATION_ID=

Generating github app private key

• Go to github app page
• Click on "generate a private key" in the Private keys section
• Save the .pem file appropriately
• Generate a base64 representation of the item using base64 -w 0 -i my.pem > encoded-private-key.txt or pbcopy < base64 -w 0 -i my.pem
• Copy the content into the GITHUB_APP_PRIVATE_KEY env variable

Commands

• make dev - runs docker environment with CLI and Web
  • CLI forwards DEV_CLI_SOURCE_URL to `DEV_CLI_TARGET_URL
  • Web defaults to http://localhost:3000 -- web alias to docker env
  • Default webhook endpoint is http://localhost:3000/webhook
• make cli-run source=http://my-source target=http://localhost:3000/webhook - runs only CLI environment, don't forget the source and target params
• make web-run - runs webserver defaulting to http://localhost:3000
• make web-run-docker - runs production version of webserver - defaults to http://localhost:8080
• kamal deploy or kamal setup to deploy things

For most cases, just run

make dev

If you want to debug server, just run cli separately

make cli-run source=http://my-source target=http://localhost:3000/webhook

Deploying the server

You can find a guide on the first article

Ansible

All credits for the playbooks are from https://github.com/guillaumebriday/kamal-ansible-manager

When provisioning a new droplet, just configure ansible inside infra folder and run it

Copy the inventory example file:

$ cp infra/hosts.ini.example infra/hosts.ini

Update the <host1> with your server's IP address (you can have multiple servers):

$ vim hosts.ini

Install the requirements:

$ ansible-galaxy collection install -r infra/requirements.yml
$ ansible-galaxy role install -r infra/requirements.yml

Configuring vars

Variables can be configured in the playbook.yml file. Also, you can override default variables provided in geerlingguy/ansible-role-swap to adjust the swap settings.

For instance:

  vars:
    security_autoupdate_reboot: "true"
    security_autoupdate_reboot_time: "03:00"
    swap_file_size_mb: '1024'

Running the playbook

Run the playbook:

$ ANSIBLE_HOST_KEY_CHECKING=False ansible-playbook -i infra/hosts.ini infra/playbook.yml