Background Jobs Patterns

Guidance for creating robust background jobs in Rails (Active Job + Sidekiq-friendly patterns).

1. Use ActiveJob with an adapter (Sidekiq recommended for scale)

2. Idempotency

• Ensure each job is idempotent. Use a unique job key or a database guard if needed.

3. Retry/Discard strategy

• Use retry_on StandardError, attempts: 5, wait: :exponentially_longer for transient errors
• Use discard_on ActiveRecord::RecordNotFound for permanent errors

4. Partial work and continuation

• Break large work into smaller jobs (fan-out/fan-in pattern)

5. Logging and observability

• Log job start/end and key metadata (job_id, args summary)
• Capture failures to Sentry or similar with context

6. Resource limits

• Avoid large in-memory arrays; stream or use batch processing

7. Testing

• Use inline adapter in tests and assert perform_enqueued_jobs or perform_enqueued_jobs with block

Example job skeleton:

class SyncUserJob < ApplicationJob queue_as :default retry_on Net::OpenTimeout, attempts: 3, wait: :exponentially_longer discard_on ActiveRecord::RecordNotFound

def perform(user_id) user = User.find(user_id) ExternalService.sync(user) rescue StandardError => e Rails.logger.error("SyncUserJob failed: #{e.message}") raise end end