Next steps
Where to go once a worker is draining a real queue — the guides, modules and internals worth reading next, and when.
Where to go once a worker is draining a real queue — the guides, modules and internals worth reading next, and when.
You have a producer, a worker, two queues, a retry policy that reflects what you actually expect to go wrong, and a plan for deploying it. That is the whole core. Everything below is optional, and the best time to read any of it is when you hit the problem it solves — not before.
| The problem | Read |
|---|---|
| "This job should run every night at 3am." | Scheduling — cron, delays, priorities, expiry. |
| "The user closed the tab; stop the job." | Cancellation — cooperative cancel and progress. |
| "This third-party API allows 100 requests a minute." | Rate limiting and concurrency caps. |
| "The same signup got queued twice." | Idempotency — unique keys and dedup at enqueue. |
| "The job charges a card, then fails after." | Durable steps — checkpoint inside one job so a retry replays instead of repeating. |
| "Step B needs step A's output, and C needs both." | Workflows — pipelines as a DAG, with fan-out, gates and sagas. |
| "The task needs a database handle I can't serialize." | Dependency Injection — resources, proxies, interception. |
| "I want my metrics in the dashboard I already have." | Events and middleware. |
| "How do I test code that enqueues?" | Testing — run tasks inline, or through a real worker in your suite. |
| "I need to look at the queue from a terminal, not a browser." | CLI usage — workers, stats, pause/resume, dead letters. |
| "The queue is busy at 9am and idle at 3am." | Autoscaling — scale workers from queue depth. |
Architecture documents the Rust engine underneath every SDK. It is not background reading; it is what you go to when a behaviour surprises you and the guide only told you what to type. Four places it earns its keep:
Complete programs rather than fragments, each solving one realistic problem: Examples.
If this track was your second task queue rather than your first, the comparison covers where flexiq differs by design, and the migration guide maps the concepts across. Questions that come up often are collected in the FAQ.