Kaal::Roda

Roda integration gem for Kaal.

kaal-roda depends on:

  • kaal
  • roda

It owns the Roda integration surface:

  • explicit boot wiring for Roda apps
  • backend wiring for memory, redis, SQLite, PostgreSQL, and MySQL
  • scheduler file boot loading relative to the Roda app root
  • opt-in scheduler startup and shutdown helpers
  • Roda-specific test coverage and a dummy app

Install

gem 'kaal-roda'
gem 'redis'   # for redis
gem 'sqlite3' # or pg / mysql2 for SQL

If you use SQL persistence, create the Kaal tables using Sequel migrations. kaal exposes the Sequel migration templates for:

  • SQLite: kaal_dispatches, kaal_locks, kaal_definitions
  • PostgreSQL: kaal_dispatches, kaal_definitions
  • MySQL: kaal_dispatches, kaal_definitions

Your app should also provide config/scheduler.yml.

What It Provides

  • Roda-native wiring on top of the Kaal engine
  • explicit backend injection for memory and custom backends
  • redis convenience wiring when the app passes a redis client
  • automatic SQL backend selection from the Sequel adapter unless the app passes adapter:
  • explicit lifecycle helpers so web processes do not implicitly start background scheduler threads

Minimal Roda

require 'roda'
require 'kaal/roda'

class ExampleHeartbeatJob
  def self.perform(*)
    puts 'heartbeat'
  end
end

class App < Roda
  plugin :kaal

  kaal backend: Kaal::Backend::MemoryAdapter.new,
       scheduler_config_path: 'config/scheduler.yml',
       namespace: 'my-app',
       start_scheduler: false

  route do |r|
    r.root { 'ok' }
  end
end

Redis

require 'roda'
require 'redis'
require 'kaal/roda'

class App < Roda
  REDIS = Redis.new(url: ENV.fetch('REDIS_URL'))

  plugin :kaal

  kaal redis: REDIS,
       scheduler_config_path: 'config/scheduler.yml',
       namespace: 'my-app',
       start_scheduler: false
end

SQL Backends

For SQL-backed Roda apps, pass a Sequel connection:

require 'sequel'

database = Sequel.connect(ENV.fetch('DATABASE_URL'))

class App < Roda
  plugin :kaal

  kaal database: database,
       adapter: 'postgres', # optional when Sequel can infer it
       scheduler_config_path: 'config/scheduler.yml'
end

Lifecycle

kaal-roda does not auto-start the scheduler by default.

If you want the web process to run it:

Kaal::Roda.start!

To stop it explicitly:

Kaal::Roda.stop!

If you pass start_scheduler: true to kaal, the addon starts the scheduler and installs an at_exit shutdown hook for that managed scheduler instance.

Preferred deployment model:

  • run the scheduler in a dedicated process when possible
  • use web-process startup only when you intentionally want co-located scheduling

Public API

  • Kaal::Roda.register!(app, backend: nil, database: nil, redis: nil, scheduler_config_path: 'config/scheduler.yml', namespace: nil, start_scheduler: false, adapter: nil, root: nil, environment: nil)
  • Kaal::Roda.configure_backend!(backend: nil, database: nil, redis: nil, adapter: nil, configuration: Kaal.configuration)
  • Kaal::Roda.detect_backend_name(database, adapter: nil)
  • Kaal::Roda.load_scheduler_file!(root:, environment: nil)
  • Kaal::Roda.start!
  • Kaal::Roda.stop!

Development

bin/rspec-unit
bin/rspec-e2e memory
REDIS_URL=redis://127.0.0.1:6379/0 bin/rspec-e2e redis
bin/rspec-e2e sqlite
DATABASE_URL=postgres://postgres:postgres@localhost:5432/kaal_test_auto bin/rspec-e2e pg
DATABASE_URL=mysql2://root:rootROOT!1@127.0.0.1:3306/kaal_test_auto bin/rspec-e2e mysql
bin/rubocop
bin/reek