Concurrency

Nagi's async work runs on Tokio. While an HTTP request or timer is waiting, other work can proceed. Start with Async and scopes for the syntax.

SQLite runs on a dedicated thread. CPU-heavy work also runs separately so it does not occupy an async worker for a long time.

async def main() -> Result[unit, Error]:
    result = try await cpu_sum(100000)
    print(result)
    return ok(print("Done"))

cpu_sum demonstrates offloading CPU work. At most four such operations run at once. An API for submitting arbitrary user functions is not implemented.

Passing data

Mutable values are not automatically shared between tasks. Move ownership when passing a value, or use shared for read-only sharing. Borrowed views cannot be passed to another task.

Cancelling the caller does not stop CPU work that has already started. Long-running work needs to check for a stop request itself.

Task counts and HTTP connection counts measure different things. See Measurements and HTTP load tests for results, including waiting time and memory use.