0 概述
jstack 命令用于生成虚拟机当前时刻线程快照(一般称为threaddump 或者javacore 文件)。线程快照就是当前虚拟机内每一条线程正在执行的方法堆栈集合,生成线程快照的目的主要用于定位如线程间的死锁、死循环、请求外部服务时间过长等导致机器load、cpu等过高。本文主要结合top、jstack命令来实际分析cpu过高线程。
1 jstack命令
jstack [option] pid
option | 作用 |
---|---|
-F | 当正常输出请求不被响应时候,强制输出线程堆栈 |
-l | 显示锁相关信息 |
-m | 混合模式,如果调用到本地方法栈,会显示本地方法栈 |
2 实例分析
public class ThreadTest {
public static void main(String[] args) throws Exception {
Thread thread = new Thread(new Runnable() {
@Override
public void run() {
//死循环
while (true) {
}
}
});
thread.start();
//主线程不退出
Thread.sleep(10000000);
}
}
启动程序后,
1.通过jps -l 找到相应java进程,
2.然后使用top -Hp pid (可以打印出改进程中线程运行情况)从图中可以看出20018 (16进制:4e32)比较消耗cpu
3.执行jstack pid (实际工作中可能吧比较多,可以将其输出到文件中)
jstack 20003
2017-11-08 20:25:46
Full thread dump Java HotSpot(TM) 64-Bit Server VM (25.72-b15 mixed mode):
"Attach Listener" #10 daemon prio=9 os_prio=0 tid=0x00007f3870001000 nid=0x4f3c waiting on condition [0x0000000000000000]
java.lang.Thread.State: RUNNABLE
"Thread-0" #9 prio=5 os_prio=0 tid=0x00007f38b00e4800 nid=0x4e32 runnable [0x00007f3892eab000]
java.lang.Thread.State: RUNNABLE
at ThreadTest$1.run(ThreadTest.java:7)
at java.lang.Thread.run(Thread.java:745)
"Service Thread" #8 daemon prio=9 os_prio=0 tid=0x00007f38b00c7000 nid=0x4e30 runnable [0x0000000000000000]
java.lang.Thread.State: RUNNABLE
"C1 CompilerThread2" #7 daemon prio=9 os_prio=0 tid=0x00007f38b00bc000 nid=0x4e2f waiting on condition [0x0000000000000000]
java.lang.Thread.State: RUNNABLE
"C2 CompilerThread1" #6 daemon prio=9 os_prio=0 tid=0x00007f38b00ba000 nid=0x4e2e waiting on condition [0x0000000000000000]
java.lang.Thread.State: RUNNABLE
"C2 CompilerThread0" #5 daemon prio=9 os_prio=0 tid=0x00007f38b00b7000 nid=0x4e2d waiting on condition [0x0000000000000000]
java.lang.Thread.State: RUNNABLE
"Signal Dispatcher" #4 daemon prio=9 os_prio=0 tid=0x00007f38b00b5800 nid=0x4e2c runnable [0x0000000000000000]
java.lang.Thread.State: RUNNABLE
"Finalizer" #3 daemon prio=8 os_prio=0 tid=0x00007f38b0083000 nid=0x4e2b in Object.wait() [0x00007f38935b2000]
java.lang.Thread.State: WAITING (on object monitor)
at java.lang.Object.wait(Native Method)
- waiting on <0x00000000d6f88ee0> (a java.lang.ref.ReferenceQueue$Lock)
at java.lang.ref.ReferenceQueue.remove(ReferenceQueue.java:143)
- locked <0x00000000d6f88ee0> (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)
"Reference Handler" #2 daemon prio=10 os_prio=0 tid=0x00007f38b007e000 nid=0x4e2a in Object.wait() [0x00007f38936b3000]
java.lang.Thread.State: WAITING (on object monitor)
at java.lang.Object.wait(Native Method)
- waiting on <0x00000000d6f86b50> (a java.lang.ref.Reference$Lock)
at java.lang.Object.wait(Object.java:502)
at java.lang.ref.Reference.tryHandlePending(Reference.java:191)
- locked <0x00000000d6f86b50> (a java.lang.ref.Reference$Lock)
at java.lang.ref.Reference$ReferenceHandler.run(Reference.java:153)
"main" #1 prio=5 os_prio=0 tid=0x00007f38b0009000 nid=0x4e24 waiting on condition [0x00007f38b5e12000]
java.lang.Thread.State: TIMED_WAITING (sleeping)
at java.lang.Thread.sleep(Native Method)
at ThreadTest.main(ThreadTest.java:14)
"VM Thread" os_prio=0 tid=0x00007f38b0076800 nid=0x4e29 runnable
"GC task thread#0 (ParallelGC)" os_prio=0 tid=0x00007f38b001e000 nid=0x4e25 runnable
"GC task thread#1 (ParallelGC)" os_prio=0 tid=0x00007f38b001f800 nid=0x4e26 runnable
"GC task thread#2 (ParallelGC)" os_prio=0 tid=0x00007f38b0021800 nid=0x4e27 runnable
"GC task thread#3 (ParallelGC)" os_prio=0 tid=0x00007f38b0023000 nid=0x4e28 runnable
"VM Periodic Task Thread" os_prio=0 tid=0x00007f38b00c9800 nid=0x4e31 waiting on condition
JNI global references: 6
搜索对应线程Id(4e32),就可以找到对应代码区域
"Thread-0" #9 prio=5 os_prio=0 tid=0x00007f38b00e4800 nid=0x4e32 runnable [0x00007f3892eab000]
java.lang.Thread.State: RUNNABLE
at ThreadTest$1.run(ThreadTest.java:7)
at java.lang.Thread.run(Thread.java:745)
参考文献:
[1] 深入理解java 虚拟机(第二版),周志明著