我正在努力弄清楚如何正确地使用Java的执行器。我意识到向ExecutorService提交任务有它自己的开销。然而,我惊讶地看到它是如此之高。
我的程序需要以尽可能低的延迟处理大量数据(股票市场数据)。大多数计算都是相当简单的算术操作。
我试着测试一些非常简单的东西:"Math.random() * Math.random()“
最简单的测试是在一个简单的循环中运行这个计算。第二个测试在匿名Runnable中执行相同的计算(这应该是用来度量创建新对象的成本)。第三个测试将Runnable传递给ExecutorService (这是用于度量引入执行器的成本)。
我在我的小型笔记本电脑(2个cpus,1.5g内存)上运行了这些测试:
(in milliseconds)
simpleCompuation:47
computationWithObjCreation:62
computationWithObjCreationAndExecutors:422(大约四分中有一次,前两个数最后相等)
注意,执行程序比在单个线程上执行花费的时间要长得多。对于1到8之间的线程池大小,这些数字大致相同。
问:我是不是遗漏了一些显而易见的东西,或者这些结果是预期的?这些结果告诉我,我传递给执行者的任何任务都必须做一些非平凡的计算。如果我正在处理数百万条消息,并且需要对每条消息执行非常简单(廉价)的转换,那么我可能仍然无法使用executors...trying将计算分散到多个CPU上,结果可能比在一个线程中执行它们要花费更多。设计决策比我原先想象的要复杂得多。有什么想法吗?
import java.util.concurrent.ExecutorService;
import java.util.concurrent.Executors;
import java.util.concurrent.TimeUnit;
public class ExecServicePerformance {
private static int count = 100000;
public static void main(String[] args) throws InterruptedException {
//warmup
simpleCompuation();
computationWithObjCreation();
computationWithObjCreationAndExecutors();
long start = System.currentTimeMillis();
simpleCompuation();
long stop = System.currentTimeMillis();
System.out.println("simpleCompuation:"+(stop-start));
start = System.currentTimeMillis();
computationWithObjCreation();
stop = System.currentTimeMillis();
System.out.println("computationWithObjCreation:"+(stop-start));
start = System.currentTimeMillis();
computationWithObjCreationAndExecutors();
stop = System.currentTimeMillis();
System.out.println("computationWithObjCreationAndExecutors:"+(stop-start));
}
private static void computationWithObjCreation() {
for(int i=0;i<count;i++){
new Runnable(){
@Override
public void run() {
double x = Math.random()*Math.random();
}
}.run();
}
}
private static void simpleCompuation() {
for(int i=0;i<count;i++){
double x = Math.random()*Math.random();
}
}
private static void computationWithObjCreationAndExecutors()
throws InterruptedException {
ExecutorService es = Executors.newFixedThreadPool(1);
for(int i=0;i<count;i++){
es.submit(new Runnable() {
@Override
public void run() {
double x = Math.random()*Math.random();
}
});
}
es.shutdown();
es.awaitTermination(10, TimeUnit.SECONDS);
}
}发布于 2022-01-22 19:49:17
如果它对其他人有用的话,下面是一个现实场景的测试结果--在三星安卓设备上反复使用ExecutorService直到所有任务结束。
Simple computation (MS): 102
Use threads (MS): 31049
Use ExecutorService (MS): 257代码:
ExecutorService executorService = Executors.newFixedThreadPool(1);
int count = 100000;
//Simple computation
Instant instant = Instant.now();
for (int i = 0; i < count; i++) {
double x = Math.random() * Math.random();
}
Duration duration = Duration.between(instant, Instant.now());
Log.d("ExecutorPerformanceTest", "Simple computation (MS): " + duration.toMillis());
//Use threads
instant = Instant.now();
for (int i = 0; i < count; i++) {
new Thread(() -> {
double x = Math.random() * Math.random();
}
).start();
}
duration = Duration.between(instant, Instant.now());
Log.d("ExecutorPerformanceTest", "Use threads (MS): " + duration.toMillis());
//Use ExecutorService
instant = Instant.now();
for (int i = 0; i < count; i++) {
executorService.execute(() -> {
double x = Math.random() * Math.random();
}
);
}
duration = Duration.between(instant, Instant.now());
Log.d("ExecutorPerformanceTest", "Use ExecutorService (MS): " + duration.toMillis());https://stackoverflow.com/questions/1647990
复制相似问题