Skip to content
This repository has been archived by the owner on Jan 26, 2022. It is now read-only.

floriandejonckheere/code-red

Repository files navigation

Code Red

Continuous Integration Continuous Deployment

Release Deployment

Code Red is a simple project management app, using a graph-based approach to task and resource management. This app was built for the Build on Redis Hackathon 2021.

A demo server is publicly available at https://codered.pm/, the database resets to sample data every 15 minutes.

Screenshot

What it does

Code Red is a simple app that allows you to manage and visualize your heavily interdependent project using a directed graph.

A project consists of a single graph, containing many tasks. A task can be an idea, goal, epic, feature, simple task or bug. It has a status as well: to do, in progress, in review or done.

Tasks can be linked to each other using a generic link (related to), or a specific relationship (blocked by, child of).

How it works

Architecture

The application is conceived as a single Ruby on Rails monolith. For persistence, the application stores its data both relationally and in a graph: the former using PostgreSQL, the latter in RedisGraph, a Redis module by RedisLabs. Administrative data (such as users and projects) is stored relationally, while storage of tasks and the relationships between them is delegated to the graph storage.

The graph storage layer features a small custom built DSL, that translates method calls in the style of ActiveRecord into RedisGraph queries.

The web app is plain HTML sprinkled with some JavaScript (Stimulus for interactivity and D3.js/cola.js for graph visualization). The HTML is rendered server-side before being sent to the client. This means that instead of using JSON to transfer data between the server and client, HTML is sent and only the part of the DOM that changes, is replaced. Please refer to Turbo and Stimulus documentation for more information.

Graph data is fetched from a JSON endpoint using D3's JSON plugin. In order to keep the application fast and snappy, Hotwire is used as a framework.

Finally, the UI is built using TailwindCSS, Heroicons, Collecticons and QuillJS for the rich text editor.

Project

Projects are stored relationally in PostgreSQL. A project has the following properties:

  • id: User identifier
  • user: Owner of the project
  • name: Human readable name of the project

The project is linked to a graph (using id as graph name) A graph has many tasks.

Task

Tasks are stored as nodes in Redis Graph. A task is a graph node and has the following properties:

  • title: Task title
  • description: Rich text, multiline description of the task
  • deadline: Date
  • status: One of Todo, In Progress, Review or Done
  • type: One of Idea, Goal, Epic, Feature, Task or Bug - user: Assignee

A task can be linked to many other tasks, by relationships.

Relationship

Relationships are stored as edges in Redis Graph. A relationship is a directed graph edge and has the following properties:

  • from: Source node
  • type: One of Blocked By, Child Of, Related To
  • to: Destination node

Relationships are stored as directed edges, but in the interface both directions are rendered. For example, if Task A is blocked by Task B, Task B will be shown as "blocks Task A". It is also possible to add relationships in both directions.

Fetch tasks

A graph has many tasks, which are fetched using the following Redis Graph query:

"GRAPH.QUERY" "055616f0-a130-42b1-a3fd-81b7c8a3ef1b" "MATCH (n:Task) RETURN n" "--compact"

Create/update task

A task is created and updated with all its properties using the following query:

"GRAPH.QUERY" "055616f0-a130-42b1-a3fd-81b7c8a3ef1b" "MERGE (n:Task {id: 'f5ec1f25-0cee-49d0-9a85-1043f04ea845'}) SET n.created_at = '2021-05-15 10:20:28 UTC', n.updated_at = '2021-05-15 10:20:28 UTC', n.graph = '#<Graph name=055616f0-a130-42b1-a3fd-81b7c8a3ef1b>', n.id = 'f5ec1f25-0cee-49d0-9a85-1043f04ea845', n.title = 'Submit hackathon app', n.description = '<p>Description of my task</p>', n.deadline = '2021-05-15', n.status = 'todo', n.type = 'task', n.user_id = '25714246-be92-4d96-b1ce-cbb57aaf4747'" "--compact"

Delete task

A task is deleted using the following query:

"GRAPH.QUERY" "055616f0-a130-42b1-a3fd-81b7c8a3ef1b" "MATCH (n:Task {id: 'f5ec1f25-0cee-49d0-9a85-1043f04ea845'}) DELETE n" "--compact"

Fetch relationship

A task's related nodes are always queried based on relationship type. The related tasks are fetched using the following query:

"GRAPH.QUERY" "055616f0-a130-42b1-a3fd-81b7c8a3ef1b" "MATCH (n:Task {id: 'c9bc52a0-c436-499c-954c-da40e82f50b2'}) -[r:blocked_by]-> (m:Task) RETURN n, m, type(r) AS t" "--compact"

Add relationship

Two tasks are linked to each other using the following query:

"GRAPH.QUERY" "055616f0-a130-42b1-a3fd-81b7c8a3ef1b" "MATCH (n:Task {id: '1ad21814-69d7-47d0-a7bb-de678b86c653'}), (m:Task {id: '07427e6b-7bba-44e4-b967-8fb5ca098053'}) MERGE (n) -[r:blocked_by]-> (m)" "--compact"

Delete relationship

Two tasks are unlinked from each other using the following query:

"GRAPH.QUERY" "055616f0-a130-42b1-a3fd-81b7c8a3ef1b" "MATCH (n:Task {id: '1ad21814-69d7-47d0-a7bb-de678b86c653'}) -[r:related_to]-> (m:Task) DELETE r" "--compact"

DSL

A small Domain Specific Language was built to accommodate and simplify graph persistence. It aims at providing a small but robust interface that should feel familiar to developers used to ActiveRecord's API. The main class implementing this construction can be found at app/graph/dsl.rb.

Example of a query:

query = graph
  .match(:n, from.class.name, id: from.id)
  .to(:r, type)
  .match(:m, to.class.name)
  .delete(:r)

query.to_cypher

# => "MATCH (n:Task {id: 'c9bc52a0-c436-499c-954c-da40e82f50b2'}) -[r:blocked_by]-> (m:Task) DELETE r"

query.execute

# => []

Setup

First, ensure you have a working Docker environment.

Pull the images and start the containers:

docker-compose up -d

Compile the frontend code:

docker-compose run --rm app bin/webpack

Set up the PostgreSQL database:

docker-compose exec app rails db:setup

Load sample data into the PostgreSQL and Redis databases:

docker-compose exec app rails database:seed

The application should now be available at http://localhost:3000.

Development

Use the bin/update script to update your development environment dependencies.

If you want to enable faster compilation of assets, run Webpack dev server in the same container as the Rails server:

docker-compose exec app bin/webpack-bin-server

Debugging

To debug the server component in your IDE, start the debug instead of the app container, and connect to localhost:1234.

Testing

Run the test suite:

rspec

Secrets

Repository secrets

Github secrets for release:

Environment secrets

Github secrets for continuous deployment (process):

  • DOCKER_TOKEN (needed for Github Container Registry)

  • GANDIV5_API_KEY (needed for Let's Encrypt integration)

  • SECRET_KEY_BASE

  • SSH_HOST (deployment host)

  • SSH_USER (deployment user)

  • SSH_KEY (private key)

  • SSH_HOST_KEY (host public key)

Releasing

Update the changelog and bump the version in lib/code_red/version.rb. Create a tag for the version and push it to Github. A Docker image will automatically be built and pushed to the registry.

nano lib/code_red/version.rb
git add lib/code_red/version.rb
git commit -m "Bump version to v1.0.0"
git tag v1.0.0
git push origin master
git push origin v1.0.0

License

See LICENSE.md.