Disruptor

This library is a low latency, inter-thread communication library written in Rust.

It's heavily inspired by the brilliant Disruptor library from LMAX.

Getting Started

Add the following to your Cargo.toml file:

disruptor = "0.5.0"

To read details of how to use the library, check out the documentation on docs.rs/disruptor.

Here's a minimal example:

use disruptor::Builder;
use disruptor::BusySpin;

// The event on the ring buffer.
struct Event {
    price: f64
}

fn main() {
    // Factory closure for initializing events.
    let factory = || { Event { price: 0.0 }};

    // Closure for processing events.
    let processor = |e: &Event, sequence: i64, end_of_batch: bool| {
        // Your processing logic here.
    };

    let size = 64;
    let mut producer = Builder::new(size, factory, processor, BusySpin)
        .create_with_single_producer();
    // Publish into the Disruptor.
    for i in 0..10 {
        producer.publish(|e| {
            e.price = i as f64;
        });
    }
}// At this point, the Producer instance goes out of scope and when the
 // processor is done handling all events then the Disruptor is dropped
 // as well.

Performance

The SPSC Disruptor variant has been benchmarked and compared to Crossbeam. See the code in the benches/spsc.rs file.

The results below are gathered from running the benchmarks on a 2016 Macbook Pro running a 2,6 GHz Quad-Core Intel Core i7. So on a modern Intel Xeon the numbers should be even better. Furthermore, it's not possible to isolate cores on Mac and pin threads which would produce even more stable results. This is future work.

If you have any suggestions to improving the benchmarks, please feel free to open an issue.

To provide a somewhat realistic benchmark not only burst of different sizes are considered but also variable pauses between bursts: 0 ms, 1 ms and 10 ms.

The latencies below are the median latency per element.

No Pause Between Bursts

Latency:

Throughput:

1 ms Pause Between Bursts

Latency:

Throughput:

10 ms Pause Between Bursts

Latency:

Throughput:

Conclusion

There's clearly a difference between the Disruptor and the Crossbeam libs. However, this is not because the Crossbeam library is not a great piece of software. It is. The Disruptor trades CPU and memory resources for lower latency and higher throughput and that is why it's able to achieve these results.

Both libraries greatly improves as the burst size goes up but the Disruptor's performance is more resilient to the pauses between bursts which is one of the design goals.

Features

• Single Producer Single Consumer (SPSC). See roadmap for MPMC.
• Multi Producer Single Consumer (MPSC).
• Low-latency.
• Busy-spin wait strategies.
• Batch consumption of events.
• Thread affinity can be set for the event processor thread.

Design Choices

Everything in the library is about low-latency and this heavily influences all choices made in this library. As an example, you cannot allocate an event and move that into the ringbuffer. Instead, events are allocated on startup to ensure they are co-located in memory to increase cache coherency. (However, you can still allocate a struct on the heap and move ownership to a field in the event on the Ringbuffer. As long as you realize that this can add latency, because the struct is allocated by one thread and dropped by another. Hence, there's synchronization happening in the allocator.)

Related Work

There are multiple other Rust projects that mimic the LMAX Disruptor library:

1. Turbine
2. Disrustor

A key feature that this library supports is multiple producers from different threads that neither of the above libraries support (at the time of writing).

Roadmap

1. Support for multiple consumers/event processor threads including interdependencies.
2. Verify correctness with Miri.
3. Add a Sleeping Wait Strategy.
4. Support for batch publication.