https://github.com/oliverswitzer/balloon-bot

A Slack bot that's meant to help let engineers in your organization know when main is broken.
https://github.com/oliverswitzer/balloon-bot

Last synced: about 2 months ago
JSON representation

A Slack bot that's meant to help let engineers in your organization know when main is broken.

• Host: GitHub
• URL: https://github.com/oliverswitzer/balloon-bot
• Owner: oliverswitzer
• License: mit
• Created: 2018-12-14T21:19:18.000Z (over 6 years ago)
• Default Branch: master
• Last Pushed: 2023-11-15T18:17:53.000Z (over 1 year ago)
• Last Synced: 2025-03-22T03:31:40.948Z (2 months ago)
• Language: Ruby
• Homepage:
• Size: 37.2 MB
• Stars: 4
• Watchers: 2
• Forks: 0
• Open Issues: 55
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
[![CircleCI](https://circleci.com/gh/oliverswitzer/balloon-bot.svg?style=shield)](https://circleci.com/gh/oliverswitzer/balloon-bot)

# Balloon Bot

This Slack bot is meant to help let engineers in your organization know when main is broken. It also will help gather information for post-mortems.

# How it works

1. When something breaks, tell balloon bot!

![Hold deployments example](images/hold-deploys.png)

2. Balloon bot will set the channel topic

![Set the channel topic](images/channel-topic.png)

3. As well as make sure to set failing github statuses on all open pull requests

![It sets github status on all open pull requests](images/pr-status.png)

4. When you've fixed the issue, tell balloon bot and all will continue as if nothing ever happened.

![You can resume deployments too!](images/back-to-green.png)

# Development

To run the bot locally, you first must setup your API token. 

1. Create a `.env` file that looks like `.env.sample` and fill it out

1. Install foreman `gem install foreman`

1. Run the app with `foreman start` or `bundle exec rake server` in the root directory

### Creating and running migrations

#### Postgres database setup

**Install postgres**

`$ brew install postgresql`

`$ brew services start postgresql`

#### Create the balloonbot database

`$ cd web && bundle exec rake db:create`

**To create a new migration**

From the top-level of the directory (not `web`), run:

`$ bundle exec rake db:generate:migration[]`

e.g.

`$ bundle exec rake db:generate:migration[CreateUsersTable]`

This will output the migration into `persistence/lib/migrations/20190406001400_create_users_table.rb` (not `web/db/migrate`)

**To run migrations**

From the top-level of the directory, run:

`$ bundle exec rake db:migrate`

All migrations in `persistence/lib/migrations` will be run.

**For any other Rails rake task feel free to run these in the `web/` directory**

### Webhook behavior

If you'd like to test the webhook functionality with Github locally (marking new PRs as failing when there's an incident), you will need to use ngrok.

* `gem install ngrok`

* Start ngrok 

`nkgrok http 5000` 

foreman automatically chooses port 5000 for now. Output in terminal should look like this: 

![ngrok_how_to!](images/ngrok.png)

* Under the configured github repo's settings, add a webhook to point at the app

 

![github_webhooks!](images/github-webhooks.png)

* Select "Add webhook"

* Enter the ngrok Forwarding url (http://cca022c1.ngrok.io in the example):

`http://cca022c1.ngrok.io/pull-request`

* Select "application/json" for Content type

* Select "Let me select individual events"

* Uncheck "Pushes" and check "Pull requests"

* Click "Add Webhook"

* The app running locally should now respond to incoming github webhooks!

Note that you'll have to update this webhook in the future if you run ngrok again, since you get assigned a new `Forwarding` url each time you run ngrok. It's a bit unfortunate, but part of development :/

# Running tests

1. Create a `.env.test` file. 

- The credentials entered here will be used to run integration tests against `Clients::Github::Wrapper` on a test repo

specified by the `GITHUB_REPO` env var. 

    - Note that two specific branches must exist in this repo for these tests to succeed, as indicated by the constants

    `TEST_BRANCH_1` and `TEST_BRANCH_2` in the spec.

    - It is recommended to use a separate github repo than what you use for local development

    to avoid possible data conflicts from simultaneous test runs 

1. `bundle exec rake test`

# Deployments

1. `git push heroku master`

2. If there are migrations

```bash

$ herkoku run rake db:migrate

```

To stop the app:

`$ heroku ps:scale web=0`

To start it back up again:

`$ heroku ps:scale web=1`

### Spinning up a new Heroku environment

**Database**

You'll need to configure a [postgres database on Heroku](https://www.heroku.com/postgres) for this app to use. The database connection

will be read in through the `DATABASE_URL` ENV var that Heroku sets for you.

**Buildpacks**

This app uses a node buildpack as well as a ruby buildpack when being deployed. Though the app is in ruby, nodejs

is necessary because we use webpack for our front-end assets.

You will need to specify which buildpacks for Heroku to use when setting up a new environment. 

You can do this with:

```bash

heroku buildpacks:clear

heroku buildpacks:set heroku/nodejs

heroku buildpacks:add heroku/ruby --index 2

```

--index 2 specifies that the ruby buildpack should come after the nodejs buildpack.