DOCKER SETUP

Build

This should only have to be done once, or whenever the Gemfile is updated.

docker-compose build

Create Containers

docker-compose up

DNA Admin should now be available at localhost:8080, but it probably needs to be set up first.

Set up system

In a new terminal run:

docker-compose exec web rails db:create db:schema:load db:migrate &&
docker-compose exec -e ADMIN_EMAIL=spree@example.com -e ADMIN_PASSWORD=spree123 web rails db:seed &&
docker-compose exec web rails spree_sample:load &&
docker-compose restart

The Default User

email: spree@example.com password: spree123

Reset DB

This will reset the existing database back to blank.

docker-compose exec web rails db:reset railties:install:migrations db:migrate db:seed spree_sample:load

You could also blow away all the DB files. WARNING! You'll have to start the install over again if you do this.

sudo rm -rf tmp/db

Extensions

The system uses 3 spree extensions

• spree_reffiliate (Thanks @Gaurav2728) github
• spree_static_content github
• spree_digital github
• spree_promo_users_codes github

Each one is installed after spree, with it's own migrations generated using a specific bundle exec rails g command, which can be found on the README of the github page for each project. This only needs to be done once after spree is installed or upgraded.

Scripts

1. Generate Affiliate Codes for Existing Users: bundle exec rake reffiliate:generate
2. Create or reset a New Admin User: docker-compose exec web rails spree_auth:admin:create

Deploy

This uses heroku ruby buildpack on the heroku-20 stack. The master branch on github is hooked in to the deployment.

Git: https://github.com/1instinct/dna-admin

Steps:

• Create pipelines on Heroku
• Add Github repos
• Create apps
• Add PG add-on
• heroku config:set -a app-name
• heroku stack:set heroku-20 -a app-name
• heroku run rails db:seed -a app-name
• heroku run rails db:schema:load db:migrate -a app-name
• heroku run rails spree_auth:admin:create -a app-name
• heroku run rails spree_sample:load -a app-name

Testing Production Settings

To test the production settings locally (used to test things like the S3 buckets for active storage), you need to set environment variables directly in the docker-compose.yml file. The production environment is configured to NOT use .env files.

To do this, apply the following patch to docker-compose.yml (after filling in real values for the keys and bucket name):

--- docker-compose.yml.orig     2021-06-02 10:50:59.011383071 -0400
+++ docker-compose.yml  2021-06-02 10:51:03.267414021 -0400
@@ -16,4 +16,10 @@
     depends_on:
       - db
     environment:
-      DATABASE_URL: postgres://postgres:password@db:5432/dna_admin_development
+      DATABASE_URL: postgres://postgres:password@db:5432/dna_admin_production
+      RAILS_ENV: production
+      AWS_REGION_NAME: us-west-1
+      AWS_BUCKET_NAME:
+      AWS_ACCESS_KEY_ID:
+      AWS_SECRET_ACCESS_KEY:
+      RAILS_SERVE_STATIC_FILES: 1

After building and starting the container, you will need to build the assets in the local container with:

docker-compose exec web rails assets:precompile
docker-compose restart

Keeping Your Code Updated

When there are lots of active changes occuring on this repo, make sure to regularly:

1. Commit (or stash) your local changes on your branch
2. git fetch origin
3. git checkout main
4. git pull origin main
5. git checkout <your_branch>
6. git rebase origin/main
7. Fix merge conflicts (if any)
8. git add .
9. git commit
10. git rebase --continue

Done! …now you will be up-to-date with latest code. Do this before you submit your PR, and you can be sure it will be a clean merge.

TODO

Other things we may need to cover:

1. Ruby version

2. System dependencies

3. Configuration

4. Database creation

5. Database initialization

6. How to run the test suite

7. Services (job queues, cache servers, search engines, etc.)

8. Deployment instructions