richardstartin commented on issue #7866:
URL: https://github.com/apache/pinot/issues/7866#issuecomment-985996820
In fact there's a simpler problem with the benchmark (once you prevent dead
code elimination you should see ~2-5ns for `benchmarkThreadCpuTimer`) but you
need to call `ThreadTimer.setThreadCpuTimeMeasurementEnabled(true)` to actually
measure anything:
```java
@BenchmarkMode(Mode.AverageTime)
@OutputTimeUnit(TimeUnit.NANOSECONDS)
@Warmup(iterations = 5, time = 1)
@Measurement(iterations = 5, time = 1)
@Fork(1)
@State(Scope.Benchmark)
public class BenchmarkTimer {
public static void main(String[] args)
throws Exception {
ChainedOptionsBuilder opt = new
OptionsBuilder().include(BenchmarkTimer.class.getSimpleName());
new Runner(opt.build()).run();
}
@Setup(Level.Trial)
public void setup() {
ThreadTimer.setThreadCpuTimeMeasurementEnabled(true);
}
@Benchmark
public long benchmarkThreadCpuTimer() {
ThreadTimer threadTimer = new ThreadTimer();
return threadTimer.getThreadTimeNs();
}
@Benchmark
public long benchmarkSystemCurrentTimeMillis() {
long startWallClockTimeMs = System.currentTimeMillis();
long totalWallClockTimeMs = System.currentTimeMillis() -
startWallClockTimeMs;
return TimeUnit.MILLISECONDS.toNanos(totalWallClockTimeMs);
}
@Benchmark
public long benchmarkSystemNanoTime() {
long startWallClockTimeNs = System.nanoTime();
return System.nanoTime() - startWallClockTimeNs;
}
}
```
I ran this on my MacBook Pro
```
Benchmark Mode Cnt Score
Error Units
BenchmarkTimer.benchmarkSystemCurrentTimeMillis avgt 5 52.619 ±
4.729 ns/op
BenchmarkTimer.benchmarkSystemNanoTime avgt 5 61.667 ±
6.172 ns/op
BenchmarkTimer.benchmarkThreadCpuTimer avgt 5 2504.818 ±
181.180 ns/op
```
Again, if you want to see this really blow up, run it on Xen.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
