Recommended way to implement a Channel consumer with async handlers & backpressure?

Hey everyone

I’m working on a microservice (DDD style) where aggregates communicate through an in-memory Channel (System.Threading.Channels).
Producers push domain messages/events into it, and a background consumer runs handlers for each message.

I want to keep backpressure (so producers block if the consumer is behind) but still handle multiple messages asynchronously — without serializing everything or flooding the system.

Here’s the trade-off I’m running into:

await foreach (var msg in reader.ReadAllAsync())
{
    await HandleAsync(msg); // preserves backpressure but no parallelism
}


await foreach (var msg in reader.ReadAllAsync())
{
    _ = Task.Run(() => HandleAsync(msg)); // loses backpressure
}

await foreach (var msg in reader.ReadAllAsync())
{
    await HandleAsync(msg); // preserves backpressure but no parallelism
}


await foreach (var msg in reader.ReadAllAsync())
{
    _ = Task.Run(() => HandleAsync(msg)); // loses backpressure
}

C#•4mo ago•

48 replies

Merlin

Recommended way to implement a Channel consumer with async handlers & backpressure?

Hey everyone

await foreach (var msg in reader.ReadAllAsync())
{
    await HandleAsync(msg); // preserves backpressure but no parallelism
}


await foreach (var msg in reader.ReadAllAsync())
{
    _ = Task.Run(() => HandleAsync(msg)); // loses backpressure
}

await foreach (var msg in reader.ReadAllAsync())
{
    await HandleAsync(msg); // preserves backpressure but no parallelism
}


await foreach (var msg in reader.ReadAllAsync())
{
    _ = Task.Run(() => HandleAsync(msg)); // loses backpressure
}

Recommended way to implement a Channel consumer with async handlers & backpressure?

Similar Threads

Recommended way to implement a Channel consumer with async handlers & backpressure?

Similar Threads

Similar Threads

Similar Threads