Background Jobs and Scheduling
Move long-running or retry-prone work out of request/response paths.
Recommended approach
- Use queue-based workers for asynchronous processing.
- Use schedulers for periodic jobs with explicit ownership.
- Keep jobs idempotent and observable.
Alternatives and when to choose them
- Cron-only approach for very small systems.
- Workflow engines for complex multi-step job orchestration.
Implementation checklist
- Define queue names and retry policy per job type.
- Add dead-letter handling and alerting.
- Add job-level tracing/log correlation IDs.
Common pitfalls
- Hidden side effects making retries unsafe.
- Scheduling logic spread across multiple apps without ownership.