Skip to content
143 changes: 143 additions & 0 deletions benchmarks/src/jmh/kotlin/benchmarks/ChannelBenchmark.kt
Original file line number Diff line number Diff line change
@@ -0,0 +1,143 @@
package benchmarks

import kotlinx.coroutines.*
import kotlinx.coroutines.channels.*
import org.openjdk.jmh.annotations.*
import java.util.concurrent.*

@Warmup(iterations = 7, time = 1)
@Measurement(iterations = 10, time = 1)
@BenchmarkMode(Mode.AverageTime)
@OutputTimeUnit(TimeUnit.NANOSECONDS)
@State(Scope.Benchmark)
@Fork(1)
open class ChannelBenchmark {
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Could you please elaborate what exactly you're trying to measure using these benchmarks?
Right now, it looks like "time required to create a new channel, send N messages into it (and, optionally, receive them), and then close the channel". However, I thought that initial idea was to measure the latency of sending (and receiving) a single message into the channel.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

measure the latency of sending (and receiving) a single message into the channel

I do measure that, indirectly. Do you suggest to literally only send/receive one message per benchmark? Is that reliable?

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Direct measurements are always better than indirect. If the goal is to measure send/recv timing, let's measure it.

What makes you think it will be unreliable?

Copy link
Contributor Author

@murfel murfel Nov 6, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Direct measurements are always better than indirect.

Not always. See below. Also depends on how you define "better".

If the goal is to measure send/recv timing, let's measure it.

Again, I am measuring that. My way of measuring is a valid way of measuring. Having to assert that makes me feel dismissed.

We can explore other ways to measure, for sure.

What makes you think it will be unreliable?

  1. Overhead of the measuring setup could be greater than the effect measured.
  2. Yet simplified setup might not capture a typical usage.
  3. Does not average over data structure amortization (e.g. our sent element could be the element which triggers the channel's internal data structure doubling / allocation) (or, on the contrary, the constant from amortization could be noticeable and we do in fact want to measure it)
  4. Does not average over GC

What setup did you have in mind, something like this?

@Benchmark
fun sendReceiveUnlimitedPrefilledSequential(wrapper: UnlimitedChannelWrapper, blackhole: Blackhole) =
    runBlocking {
        wrapper.channel.send(42)
        blackhole.consume(wrapper.channel.receive())
    }
ChannelBenchmark.sendReceiveUnlimitedPrefilledSequential         0          0  avgt   10  53.959 ±  0.168  ns/op
ChannelBenchmark.sendReceiveUnlimitedPrefilledSequential         0    1000000  avgt   10  60.069 ±  1.345  ns/op
ChannelBenchmark.sendReceiveUnlimitedPrefilledSequential         0  100000000  avgt   10  71.457 ± 13.101  ns/op

Or this? (no suspension, trySend/tryReceive)

@Benchmark
fun sendReceiveUnlimitedPrefilledSequentialNoSuspension(wrapper: UnlimitedChannelWrapper, blackhole: Blackhole) {
    wrapper.channel.trySend(42)
    blackhole.consume(wrapper.channel.tryReceive().getOrThrow())
}
Benchmark                                                  (count)  (prefill)  Mode  Cnt   Score   Error  Units
ChannelBenchmark.sendReceiveUnlimitedPrefilledNoSuspension        0          0  avgt   10  10.619 ± 0.270  ns/op
ChannelBenchmark.sendReceiveUnlimitedPrefilledNoSuspension        0    1000000  avgt   10  10.859 ± 0.330  ns/op
ChannelBenchmark.sendReceiveUnlimitedPrefilledNoSuspension        0  100000000  avgt   10  17.163 ± 1.523  ns/op

Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The internal structure of the channel may be worth taking into account. For a prefilled channel with 32+ elements (32 is the default channel segment size), we can expect send and receive not to interact with one another at all, that is, the duration of send followed by receive should be roughly the sum of durations of send and receive, invoked independently. I imagine wrapper.channel.send(42) and blackhole.consume(wrapper.channel.receive()) could stay in different benchmarks without affecting the results too much.

For an empty channel, we could also try racing send and receive.

Using runBlocking in a benchmark that's only doing send doesn't seem optimal to me, I can imagine the run time getting dominated by the runBlocking machinery. I don't know what the proper way of doing this in JMH is, but I'd try a scheme like this:

internal class BenchmarkSynchronization() {
    private val state = AtomicInteger(0)
    private val benchmarkThread = Thread.currentThread()
    private val threadDoingWork = AtomicReference<Thread?>()
    
    fun awaitThreadAssignment(): Thread {
        assert(Thread.currentThread() === benchmarkThread)
        while (true) {
            val thread = threadDoingWork.get()
            if (thread != null) return thread
            LockSupport.parkNanos(Long.MAX_VALUE)
        }
    }
    
    fun awaitStartSignal() {
        threadDoingWork.set(Thread.currentThread())
        LockSupport.unpark(benchmarkThread)
        while (state.get() == 0) {
            LockSupport.parkNanos(Long.MAX_VALUE)
        }
    }
    
    fun signalFinish() {
        state.set(2)
        LockSupport.unpark(benchmarkThread)
    }

    fun runBenchmark(thread: Thread) {
        state.set(1)
        LockSupport.unpark(thread)
        while (state.get() != 2) {
            LockSupport.parkNanos(Long.MAX_VALUE)
        }
    }
}

(haven't actually tested the code). Then, the scheme would be:

// preparation
val synchronization = BenchmarkSynchronization()
GlobalScope.launch {
    synchronization.awaitStartSignal()
    try {
       // actual benchmark code here
    } finally {
        synchronization.signalFinish()
    }
}
val threadDoingWork = synchronization.awaitThreadAssignment()

// the @Benchmark itself
wrapper.synchronization.runBenchmark(wrapper.threadDoingWork)

@fzhinkin , is there a standard mechanism that encapsulates this?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

For an empty channel, we could also try racing send and receive.

?

Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Running them in parallel.

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Direct measurements are always better than indirect.

Not always. See below. Also depends on how you define "better".

I mean, if our goal is to measure a latency of a certain operation and we have facilities to do so, then it's better to do it directly (to an extent, benchmark's results are averages anyway). By doing so, we can ensure that all unnecessary setup and teardown code (like, creating a channel) won't skew results.

On the other hand, if the goal is to measure end-to-end latency, like "time to create a channel and send 100k messages over it", then sure, the current approach works for that (moreover, I don't see how to measure it otherwise).

If the goal is to measure send/recv timing, let's measure it.

Again, I am measuring that. My way of measuring is a valid way of measuring. Having to assert that makes me feel dismissed.

See the comment, above. I was under the impression that the typical use case for channel is to be used indirectly (within a flow, for example), so for channels as they are we decided to measure a latency of a single operation to see how it will be affected by potential changes in the implementation.

I'm not saying that the way you're measuring it is invalid, but if there are facilities to measure latency of a single operation (well, the send-receive pair of operations), I'm voting for using it (unless there is an evidence that such a measurement is impossible or makes no sense).

We can explore other ways to measure, for sure.

What makes you think it will be unreliable?
Overhead of the measuring setup could be greater than the effect measured.

Setup (and teardown) actions performed before (after) the whole run (or an individual iteration) should not affect measurements (as they are performed outside of the measurement scope); it will affect the measurements when performed for each benchmark function invocation.

Yet simplified setup might not capture a typical usage.

I'm not sure if sending 400MB of data is a typical usage either. ;)

Does not average over data structure amortization (e.g. our sent element could be the element which triggers the channel's internal data structure doubling / allocation) (or, on the contrary, the constant from amortization could be noticeable and we do in fact want to measure it)

The benchmark function is continuously invoked over a configured period of time (you set it to 1 second).
If we reuse the same channel in each invocation, results will average over data structure amortization.

Does not average over GC

It's easier to focus on memory footprint as it is something we control directly (how many bytes we're allocating when performing an operation), rather than on GC pauses (they are a subject to various factors).

What setup did you have in mind, something like this?

Both approaches look sane (assuming the wrapper is not recreated for every benchmark call) and we can do both.

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@dkhalanskyjb, it feels like I didn't get you, but nevertheless: JMH provides some facilities to running benchmark methods concurrently and synchronize their execution:
https://github.com/openjdk/jmh/blob/master/jmh-samples/src/main/java/org/openjdk/jmh/samples/JMHSample_15_Asymmetric.java
https://github.com/openjdk/jmh/blob/master/jmh-samples/src/main/java/org/openjdk/jmh/samples/JMHSample_17_SyncIterations.java

As of runBlocking, it would be nice to have a kx-benchmarks maintainer here, who would solve a problem with benchmarking suspend-API for us. Oh, wait... 😄

// max coroutines launched per benchmark
// to allow for true parallelism
val cores = Runtime.getRuntime().availableProcessors()

// 4 KB, 40 KB, 400 KB, 4 MB, 40 MB, 400 MB
@Param(value = ["1000", "10000", "100000", "1000000", "10000000", "100000000"])
var count: Int = 0

// 1. Preallocate.
// 2. Different values to avoid helping the cache.
val list = List(100000000) { it }

@State(Scope.Benchmark)
open class UnlimitedChannelWrapper {
// 0, 4 MB, 40 MB, 400 MB
@Param(value = ["0", "1000000", "10000000", "100000000"])
private var prefill = 0

lateinit var channel: Channel<Int>

val list = List(100000000) { it }

@Setup(Level.Invocation)
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Why it has to be done before every benchmark function invocation and not once per trial / iteration?

JFTR, https://github.com/openjdk/jmh/blob/2a316030b509aa9874dd6ab04e21962ac92cd634/jmh-core/src/main/java/org/openjdk/jmh/annotations/Level.java#L85

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It's a tradeoff. After each invocation, there could be a little extra items in the channel, which can accumulate with iterations. I can rewrite runSendReceive to leave channel with the same number of elements as it came in with, but then it will slightly affect the benchmark. Possibly negligible, since it's only up to 4 items each time...

fun createPrefilledChannel() {
channel = Channel(Channel.UNLIMITED)
repeat(prefill) {
channel.trySend(list[it])
}
}
}


@Benchmark
fun sendUnlimited() = runBlocking {
runSend(count, Channel.UNLIMITED)
}

@Benchmark
fun sendConflated() = runBlocking {
runSend(count, Channel.CONFLATED)
}

@Benchmark
fun sendReceiveUnlimited(wrapper: UnlimitedChannelWrapper) = runBlocking {
runSendReceive(wrapper.channel, count)
}

@Benchmark
fun sendReceiveConflated() = runBlocking(Dispatchers.Default) {
runSendReceive(Channel(Channel.CONFLATED), count)
}

@Benchmark
fun sendReceiveRendezvous() = runBlocking(Dispatchers.Default) {
// NB: Rendezvous is partly benchmarking the scheduler, not the channel alone.
// So don't trust the Rendezvous results too much.
runSendReceive(Channel(Channel.RENDEZVOUS), count)
}

@Benchmark
fun oneSenderManyReceivers(wrapper: UnlimitedChannelWrapper) = runBlocking {
runSendReceive(wrapper.channel, count, 1, cores - 1)
}

@Benchmark
fun manySendersOneReceiver(wrapper: UnlimitedChannelWrapper) = runBlocking {
runSendReceive(wrapper.channel, count, cores - 1, 1)
}

@Benchmark
fun manySendersManyReceivers(wrapper: UnlimitedChannelWrapper) = runBlocking {
runSendReceive(wrapper.channel, count, cores / 2, cores / 2)
}

private suspend fun runSend(count: Int, capacity: Int) {
val channel = Channel<Int>(capacity)
repeat(count) {
channel.send(list[it])
}
}

suspend fun <E> Channel<E>.forEach(action: (E) -> Unit) {
for (element in this) {
action(element)
}
}

// NB: not all parameter combinations make sense in general.
// E.g., for the rendezvous channel, senders should be equal to receivers.
// If they are non-equal, it's a special case of performance under contention.
private suspend inline fun runSendReceive(channel: Channel<Int>, count: Int, senders: Int = 1, receivers: Int = 1) {
//require (senders > 0 && receivers > 0)
//require (senders + receivers <= cores) // Can be used with more than num cores, but what would it measure?
// if the channel is prefilled, only receive the items that were sent by this function
val receiveAll = channel.isEmpty
// send almost `count` items, up to `senders - 1` items will not be sent (negligible)
val countPerSender = count / senders
// for prefilled channel only: up to `receivers - 1` items of the sent items will not be received
// (on top of the prefilled items which we do not aim to receive at all) (negligible)
val countPerReceiverAtLeast = countPerSender * senders / receivers
withContext(Dispatchers.Default) {
repeat(receivers) {
launch {
if (receiveAll) {
channel.forEach {
// possibly receive into the blackhole
}
} else {
repeat(countPerReceiverAtLeast) {
// possibly receive into the blackhole
channel.receive()
}
}
}
}

coroutineScope {
repeat(senders) {
launch {
repeat(countPerSender) {
channel.send(list[it])
}
}
}
}
channel.close()
}
}
}