Python Concurrency, Async, and Parallelism¶
Choose concurrency from the workload, then bound tasks, queues, time, and shared state.
flowchart LR
Work --> IO{mostly waiting?}
IO -->|async-compatible| Async[asyncio]
IO -->|blocking I/O| Threads[threads]
IO -->|CPU work| Processes[processes / native code]
Progress through safe basics, model selection, saturation behavior, and capacity governance.