POD Aggregator

The POD Aggregator project is a Ruby on Rails application that receives and transmits MARC bibliographic and holdings data from multiple institutions.

Developing

Pre-requisites

This project is tested on ruby 3 and nodeJS 16. Other versions may work but are unsupported. JavaScript package management is done via yarn.

Getting started

Pull down the code and enter the project directory:

git clone https://github.com/ivplus/aggregator.git
cd aggregator

Install dependencies and prepare the database:

bin/setup

Create a local admin user for development:

bin/rails agg:create_admin

Configurations

POD Aggregator has several configuration settings many of which are available using the config gem at config/settings.yml.

Adding data

While adding MARC data locally is perfectly fine for many development use cases, if you want to populate your local development instance with an organization and a small amount of data to get started, use the rake db:seed task:

bundle exec rake db:seed

To more broadly populate the database, use the configurable agg:seed_from_api task. This will enable fetching data from a running aggregator instance (production by default). The configuration file config/settings.yml controls what data gets pulled with the agg:seed_from_api task. Add a valid API token to the settings.yml file in the token field under marc_seed_fixtures and update the list of organizations to import data from as necessary.

Depending on the data available, this task might load a large amount of data and take a long time:

bundle exec rake agg:seed_from_api

Manually triggering normalization jobs

To generate normalized data locally you can trigger a cron job from the terminal. This will generate normalized full dumps, deletes, and deltas.

bundle exec rails runner "GenerateFullDumpJob.enqueue_all"

Testing

The continuous integration tests for POD aggregator can be run using:

bundle exec rake

Emulating production

While using a standard development setup locally, developers may experience some issues with read/write concurrency in Sqlite and job processing using the ActiveJob AsyncAdapter.

One solution to this is to temporarily switch to using PostgreSQL and Sidekiq with Redis locally for a more "production like" development environment. To do this:

Set environment variables prior to invoking the server. For now, you will need to re-export these variables in every new terminal window you open.

export SIDEKIQ_REDIS_URL=redis://localhost:6379/0
export DATABASE_URL=postgresql://user:password@localhost:5432/dbname
bin/dev  # or other commands

Deployment

Deployment is setup using Capistrano using standard Stanford Digital Library Systems and Services practices.

With Docker

Initial Setup:

1. docker compose up -d
2. docker compose exec aggapp ./bin/setup
3. Create an admin user: docker compose exec aggapp ./bin/rails agg:create_admin

Stopping and Starting:

Start the stack with:

1. docker compose up -d

Stop the stack with: 2. docker compose down

Managing Data:

Data for the redis and postgres services are persisted using docker named volumes. You can see what volumes are currently present with: docker volume ls If you want to remove a volume (e.g. to start with a fresh database or redis cache), you can do: docker volume rm aggregator_db docker volume rm aggregator_cache

Resetting Docker:

In some cases Docker container image versions may get stale and may need to be rebuilt with no cache. Attempt the steps in option 1 first before proceeding to option 2.

Option 1: Rebuild images

Step 1: Stop and remove all running docker containers for POD

docker compose down

Step 2: Rebuild docker images with no cache

docker compose build --no-cache

Step 3: Restart containers

docker compose up -d

Option 2: System prune

To completely remove all docker containers and images from the system, run the system prune command.

Step 1: System prune

Warning: The system prune command will remove all containers and images from the system.

docker system prune -a -f
docker ps -aq
docker compose pull

Optionally, to also remove all volumes, the system prune command can be run with the --volumes option. Warning: The volumes option will remove all named volumes and all local data will be deleted. The data seeds will need to be run again if the volumes are purged.

docker system prune -a -f --volumes

Logs:

To check the log output of a container run the command docker logs ${container_name} e.g.: docker logs aggapp

Development Standards

Display dates

Dates are displayed in a standard format on all UI pages across the application. The local_time library accepts a date in UTC and automatically detects the user's timezone on the client side and displays the equivalent date in the local timezone. Also there is a custom helper method datetime_display_format() that returns the standard date format. That way the date format can be managed centrally.

To display a date in the UI, use the local_time library with the datetime_display_format helper method.

<%= local_time(uploads.last.created_at, format: datetime_display_format()) %>