java线程内存占用_java – 线程占用的内存

我需要监视由我的应用程序生成的线程消耗的内存量.如果贪婪的线程消耗太多内存,那么我们的想法是采取纠正措施.我已经提到了

How much memory does my java thread take?.关于该链接的一个建议是在ThreadMXBean中使用getThreadAllocatedBytes.我使用以下作业对getThreadAllocatedBytes进行了实验.

List primes = new ArrayList();

long i = 0;

while (true) {

primes.add(++i);

if ((i % 10) == 0) {

primes.clear();

System.runFinalization();

System.gc();

}

}

我花了很长时间在四个线程上运行这个工作.虽然作业不会持续累积内存,但getThreadAllocatedBytes返回的值会不断增加,甚至不会下降一次.这意味着getThreadAllocatedBytes不会返回线程使用的堆上的实际内存量.它返回自启动以来线程在堆上分配的内存总量.我的平台详细信息如下：

Linux PG85213.egi.ericsson.com 3.5.0-030500-generic#201207211835 SMP Sat Jul 21 21:35:55 UTC 2012 x86_64 x86_64 x86_64 GNU / Linux

java版“1.7.0_45”

Java(TM)SE运行时环境(版本1.7.0_45-b18)

Java HotSpot(TM)64位服务器VM(内置24.45-b08,混合模式)

上述行为是否是getThreadAllocatedBytes所需的行为？

如果是这样,是否无法在线程使用的堆上找到有效内存.

我列出了完整的程序供参考：

package workbench;

import java.lang.management.ManagementFactory;

import com.sun.management.ThreadMXBean;

import java.util.ArrayList;

import java.util.Collections;

import java.util.List;

import java.util.concurrent.CountDownLatch;

import java.util.concurrent.Executors;

import java.util.logging.Level;

import java.util.logging.Logger;

public class AnotherWorkBench {

private static final CountDownLatch latch = new CountDownLatch(4);

static final List threadIds = Collections.synchronizedList(new ArrayList());

private void dummyJob() {

List primes = new ArrayList();

long i = 0;

while (true) {

primes.add(++i);

if ((i % 10) == 0) {

primes.clear();

//introduce sleep to prevent process hogging

try {

Thread.currentThread().sleep(2000);

} catch (InterruptedException ex) {

Logger.getLogger(AnotherWorkBench.class.getName()).log(Level.SEVERE, null, ex);

}

System.runFinalization();

System.gc();

}

}

}

private void runDummyJobs() {

Runnable dummyJob = new Runnable() {

@Override

public void run() {

threadIds.add(Thread.currentThread().getId());

latch.countDown();

dummyJob();

}

};

Runnable memoryMonitorJob = new Runnable() {

@Override

public void run() {

System.out.println(Thread.currentThread().getName() + " : Monitor thread started");

ThreadMXBean threadMxBean = (ThreadMXBean) ManagementFactory.getThreadMXBean();

threadMxBean.setThreadAllocatedMemoryEnabled(true);

while (true) {

for (Long threadId : threadIds) {

System.out.println(Thread.currentThread().getName() + " : Thread ID : " + threadId + " : memory = " + threadMxBean.getThreadAllocatedBytes(threadId) + " bytes");

}

//wait between subsequent scans

try {

System.out.println(Thread.currentThread().getName() + " : secondary sleep");

Thread.currentThread().sleep(5000);

System.out.println(Thread.currentThread().getName() + " : out of secondary sleep");

} catch (InterruptedException ex) {

Logger.getLogger(WorkBench.class.getName()).log(Level.SEVERE, null, ex);

}

}

}

};

Executors.newSingleThreadExecutor().submit(dummyJob);

Executors.newSingleThreadExecutor().submit(dummyJob);

Executors.newSingleThreadExecutor().submit(dummyJob);

Executors.newSingleThreadExecutor().submit(dummyJob);

try {

latch.await();

} catch (InterruptedException ex) {

Logger.getLogger(AnotherWorkBench.class.getName()).log(Level.SEVERE, null, ex);

}

Executors.newSingleThreadExecutor().submit(memoryMonitorJob);

}

/**

* @param args the command line arguments

*/

public static void main(String[] args) {

new AnotherWorkBench().runDummyJobs();

}

}