rails-jobs

Use ActiveJob's `retry_on` and `discard_on` methods to distinguish transient failures from permanent ones. `retry_on` automatically reschedules failed jobs with exponential backoff, while `discard_on` silently drops jobs that encounter permanent errors, keeping your queue clean and reliable.

Keep jobs thin by delegating business logic to models and services. Jobs should only handle enqueuing, argument deserialization, and calling domain methods. This separation makes jobs testable, reusable, and easier to maintain across your Rails application.

ActiveJob and Sidekiq serialize job arguments to JSON, which doesn't support all Ruby types. Pass primitive values, hashes, and IDs instead of full objects. Fetch records inside the job by ID to avoid serialization issues and handle missing records gracefully.

Yes. ActiveJob abstracts job queuing, and Sidekiq provides the reliable backend. Together they enable error handling through `retry_on` and `discard_on`, proper job naming conventions, and idempotent job design for safe retries.

Design jobs to produce the same result when run multiple times with identical arguments. Avoid side effects outside database transactions, check for existing records before creating them, and use conditional logic so repeated executions don't cause duplicates or inconsistencies.

Use `perform_now` for synchronous execution during tests or special cases, and `perform_later` for background processing. Clear naming with `_now` and `_later` semantics makes intent explicit, improves debugging, and helps teams understand when work happens asynchronously versus immediately.