如何找到占用CPU最高的是哪一个线程

最新推荐文章于 2024-05-20 14:22:21 发布

原创最新推荐文章于 2024-05-20 14:22:21 发布 · 2.1k 阅读

5 ·

CC 4.0 BY-SA版权

文章标签：

#jvm #cpu最高

jvm 专栏收录该内容

2 篇文章

订阅专栏

当遇到CPU使用率飙升的问题时，本文介绍了一种定位占用CPU最高线程的方法。通过模拟程序创建多个线程，其中有一个线程持续运行导致CPU占用增加。通过`jps`、`top -Hp`、`printf`和`jstack -l`等命令，逐步找出并分析占用CPU最高的线程1。

一、需求分析

经常听到朋友说他们线上CPU有时候突然飙升，想找到是哪个线程导致的，我相信他们肯定有自己的定位方法，今天我要分享的也仅仅是其中一种而已。

二、如何定位

为了模拟这种现象，写一个程序，先看代码

public static void main(String[] args) {
		for(int i=1;i<=3;i++) {
			new Thread() {
	            @Override
				public void run() {
					try {
						Thread.sleep(1000);
					} catch (InterruptedException e) {
						e.printStackTrace();
					}
				}
			}.start();
		}
		Thread thread = new Thread() {
			    @Override
				public void run() {
					while(true) {
						System.out.println("start foreach");
					}
				}
		};
		thread.setName("线程1");
		thread.start();
	}

有几个线程，其中，有3个线程什么都没干，仅仅是sleep了，而另外一个单独的线程我让它while(true)一直不停的输出 "start foreach"，很明显，这个线程名叫做"线程1"的线程肯定会大量占用CPU，那么在我们事先不知道的情况下，如何把它找出来呢？

首先将上面代码打成一个jar包，然后直接启动即可，会看看源源不断的输出 start foreach，

第一步：先jps看一下进程号

[root@slave ~]# jps
2942 Jps
2910 jar

第二步：使用top -Hp 进程号查看该进程下各个线程占用CPU情况

root@slave ~]# top -Hp 2910


top - 01:09:25 up 13 min,  2 users,  load average: 0.11, 0.13, 0.09
Tasks:  14 total,   0 running,  14 sleeping,   0 stopped,   0 zombie
Cpu(s):  5.1%us,  9.1%sy,  0.0%ni, 85.8%id,  0.0%wa,  0.0%hi,  0.0%si,  0.0%st
Mem:   3907472k total,  1282472k used,  2625000k free,    69612k buffers
Swap:  1048572k total,        0k used,  1048572k free,   210764k cached

   PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND                      
  2926 root      20   0 4048m 306m  11m S 11.7  8.0   0:16.61 java                         
  2910 root      20   0 4048m 306m  11m S  0.0  8.0   0:00.04 java                         
  2911 root      20   0 4048m 306m  11m S  0.0  8.0   0:00.46 java                         
  2912 root      20   0 4048m 306m  11m S  0.0  8.0   0:00.00 java                         
  2913 root      20   0 4048m 306m  11m S  0.0  8.0   0:00.00 java                         
  2914 root      20   0 4048m 306m  11m S  0.0  8.0   0:00.00 java                         
  2915 root      20   0 4048m 306m  11m S  0.0  8.0   0:00.00 java                         
  2916 root      20   0 4048m 306m  11m S  0.0  8.0   0:00.00 java                         
  2917 root      20   0 4048m 306m  11m S  0.0  8.0   0:00.00 java                         
  2918 root      20   0 4048m 306m  11m S  0.0  8.0   0:00.00 java                         
  2919 root      20   0 4048m 306m  11m S  0.0  8.0   0:00.45 java                         
  2920 root      20   0 4048m 306m  11m S  0.0  8.0   0:00.13 java                         
  2921 root      20   0 4048m 306m  11m S  0.0  8.0   0:00.00 java                         
  2922 root      20   0 4048m 306m  11m S  0.0  8.0   0:00.27 java

从上可以看到，线程号 2926的占用CPU最高。

第三步：执行 printf 0x%x 线程号得到此线程id的十六进制

[root@slave ~]# printf 0x%x 2926
0xb6e

第四步：使用jstack -l 进程号输出堆栈信息

[root@slave ~]# jstack -l 2910
2018-02-13 01:10:38
Full thread dump Java HotSpot(TM) 64-Bit Server VM (25.77-b03 mixed mode):

"Attach Listener" #14 daemon prio=9 os_prio=0 tid=0x00007f796402c000 nid=0xb9b runnable [0x0000000000000000]
   java.lang.Thread.State: RUNNABLE

   Locked ownable synchronizers:
        - None

"DestroyJavaVM" #13 prio=5 os_prio=0 tid=0x00007f799c009800 nid=0xb5f waiting on condition [0x0000000000000000]
   java.lang.Thread.State: RUNNABLE

   Locked ownable synchronizers:
        - None

"线程1" #12 prio=5 os_prio=0 tid=0x00007f799c14c800 nid=0xb6e runnable [0x00007f7983849000]
   java.lang.Thread.State: RUNNABLE
        at java.io.FileOutputStream.writeBytes(Native Method)
        at java.io.FileOutputStream.write(FileOutputStream.java:326)
        at java.io.BufferedOutputStream.flushBuffer(BufferedOutputStream.java:82)
        at java.io.BufferedOutputStream.flush(BufferedOutputStream.java:140)
        - locked <0x000000008001d830> (a java.io.BufferedOutputStream)
        at java.io.PrintStream.write(PrintStream.java:482)
        - locked <0x000000008001d780> (a java.io.PrintStream)
        at sun.nio.cs.StreamEncoder.writeBytes(StreamEncoder.java:221)
        at sun.nio.cs.StreamEncoder.implFlushBuffer(StreamEncoder.java:291)
        at sun.nio.cs.StreamEncoder.flushBuffer(StreamEncoder.java:104)
        - locked <0x000000008001da58> (a java.io.OutputStreamWriter)
        at java.io.OutputStreamWriter.flushBuffer(OutputStreamWriter.java:185)
        at java.io.PrintStream.newLine(PrintStream.java:546)
        - eliminated <0x000000008001d780> (a java.io.PrintStream)
        at java.io.PrintStream.println(PrintStream.java:807)
        - locked <0x000000008001d780> (a java.io.PrintStream)
        at com.tanjie.java.jvm.JVM如何定位消耗CPU最高的线程$2.run(JVM如何定位消耗CPU最高的线程.java:22)

   Locked ownable synchronizers:
        - None

"Service Thread" #8 daemon prio=9 os_prio=0 tid=0x00007f799c119000 nid=0xb69 runnable [0x0000000000000000]
   java.lang.Thread.State: RUNNABLE

   Locked ownable synchronizers:
        - None

"C1 CompilerThread1" #7 daemon prio=9 os_prio=0 tid=0x00007f799c0fc000 nid=0xb68 waiting on condition [0x0000000000000000]
   java.lang.Thread.State: RUNNABLE

   Locked ownable synchronizers:
        - None

"C2 CompilerThread0" #6 daemon prio=9 os_prio=0 tid=0x00007f799c0fa000 nid=0xb67 waiting on condition [0x0000000000000000]
   java.lang.Thread.State: RUNNABLE

   Locked ownable synchronizers:
        - None

"Signal Dispatcher" #5 daemon prio=9 os_prio=0 tid=0x00007f799c0f7800 nid=0xb66 runnable [0x0000000000000000]
   java.lang.Thread.State: RUNNABLE

   Locked ownable synchronizers:
        - None

"Surrogate Locker Thread (Concurrent GC)" #4 daemon prio=9 os_prio=0 tid=0x00007f799c0f6000 nid=0xb65 waiting on condition [0x0000000000000000]
   java.lang.Thread.State: RUNNABLE

   Locked ownable synchronizers:
        - None

"Finalizer" #3 daemon prio=8 os_prio=0 tid=0x00007f799c0c3800 nid=0xb64 in Object.wait() [0x00007f7983d13000]
   java.lang.Thread.State: WAITING (on object monitor)
        at java.lang.Object.wait(Native Method)
        - waiting on <0x0000000080008ee0> (a java.lang.ref.ReferenceQueue$Lock)
        at java.lang.ref.ReferenceQueue.remove(ReferenceQueue.java:143)
        - locked <0x0000000080008ee0> (a java.lang.ref.ReferenceQueue$Lock)
        at java.lang.ref.ReferenceQueue.remove(ReferenceQueue.java:164)
        at java.lang.ref.Finalizer$FinalizerThread.run(Finalizer.java:209)

   Locked ownable synchronizers:
        - None

"Reference Handler" #2 daemon prio=10 os_prio=0 tid=0x00007f799c0be800 nid=0xb63 in Object.wait() [0x00007f7983d54000]
   java.lang.Thread.State: WAITING (on object monitor)
        at java.lang.Object.wait(Native Method)
        - waiting on <0x0000000080006b50> (a java.lang.ref.Reference$Lock)
        at java.lang.Object.wait(Object.java:502)
        at java.lang.ref.Reference.tryHandlePending(Reference.java:191)
        - locked <0x0000000080006b50> (a java.lang.ref.Reference$Lock)
        at java.lang.ref.Reference$ReferenceHandler.run(Reference.java:153)

   Locked ownable synchronizers:
        - None

"VM Thread" os_prio=0 tid=0x00007f799c0b7000 nid=0xb62 runnable 

"Gang worker#0 (Parallel GC Threads)" os_prio=0 tid=0x00007f799c01a800 nid=0xb60 runnable 

"Concurrent Mark-Sweep GC Thread" os_prio=0 tid=0x00007f799c03d000 nid=0xb61 runnable 

"VM Periodic Task Thread" os_prio=0 tid=0x00007f799c124800 nid=0xb6a waiting on condition 

JNI global references: 9

然后找到nid=0xb6e的线程堆栈，即可定位问题。