虚拟机stack全分析
通过jps -lv 获取到本地的一个JVM实例进程。再通过jstack pid > thread.txt ，把stack trace输出到thread.txt文件中。
2012-08-28 23:16:39
2012-08-28 23:16:39 Full thread dump OpenJDK Client VM (19.0-b09 mixed mode, sharing): "Attach Listener" daemon prio=10 tid=0xb3da5400 nid=0x13ea waiting on condition [0x00000000] java.lang.Thread.State: RUNNABLE "Worker-15" prio=10 tid=0x09b7b400 nid=0x12f2 in Object.wait() [0xb30f9000] java.lang.Thread.State: TIMED_WAITING (on object monitor) at java.lang.Object.wait(Native Method) at org.eclipse.core.internal.jobs.WorkerPool.sleep(WorkerPool.java:188) - locked <0x6e5f87c0> (a org.eclipse.core.internal.jobs.WorkerPool) at org.eclipse.core.internal.jobs.WorkerPool.startJob(WorkerPool.java:220) at org.eclipse.core.internal.jobs.Worker.run(Worker.java:50) "Worker-12" prio=10 tid=0x09910c00 nid=0x12a0 in Object.wait() [0xb4026000] java.lang.Thread.State: TIMED_WAITING (on object monitor) at java.lang.Object.wait(Native Method) at org.eclipse.core.internal.jobs.WorkerPool.sleep(WorkerPool.java:188) - locked <0x6e5f87c0> (a org.eclipse.core.internal.jobs.WorkerPool) at org.eclipse.core.internal.jobs.WorkerPool.startJob(WorkerPool.java:220) at org.eclipse.core.internal.jobs.Worker.run(Worker.java:50) .......省略 "Low Memory Detector" daemon prio=10 tid=0x097eb000 nid=0xa9f runnable [0x00000000] java.lang.Thread.State: RUNNABLE "CompilerThread0" daemon prio=10 tid=0x097e9000 nid=0xa9e waiting on condition [0x00000000] java.lang.Thread.State: RUNNABLE "Signal Dispatcher" daemon prio=10 tid=0x097e7400 nid=0xa9d runnable [0x00000000] java.lang.Thread.State: RUNNABLE "Finalizer" daemon prio=10 tid=0x097e4000 nid=0xa9c in Object.wait() [0xb4dd1000] java.lang.Thread.State: WAITING (on object monitor) at java.lang.Object.wait(Native Method) at java.lang.ref.ReferenceQueue.remove(ReferenceQueue.java:133) - locked <0x6e2e02f8> (a java.lang.ref.ReferenceQueue$Lock) at java.lang.ref.ReferenceQueue.remove(ReferenceQueue.java:149) at java.lang.ref.Finalizer$FinalizerThread.run(Finalizer.java:177) "Reference Handler" daemon prio=10 tid=0x097df400 nid=0xa9b in Object.wait() [0xb4e22000] java.lang.Thread.State: WAITING (on object monitor) at java.lang.Object.wait(Native Method) at java.lang.Object.wait(Object.java:502) at java.lang.ref.Reference$ReferenceHandler.run(Reference.java:133) - locked <0x6e2e0098> (a java.lang.ref.Reference$Lock) "main" prio=10 tid=0x097b0400 nid=0xa99 runnable [0xbf8b9000] java.lang.Thread.State: RUNNABLE at org.eclipse.swt.internal.gtk.OS.Call(Native Method) at org.eclipse.swt.widgets.Display.sleep(Display.java:4069) at org.eclipse.ui.application.WorkbenchAdvisor.eventLoopIdle(WorkbenchAdvisor.java:364) at org.eclipse.ui.internal.ide.application.IDEWorkbenchAdvisor.eventLoopIdle(IDEWorkbenchAdvisor.java:917) at org.eclipse.ui.internal.Workbench.runEventLoop(Workbench.java:2697) at org.eclipse.ui.internal.Workbench.runUI(Workbench.java:2660) at org.eclipse.ui.internal.Workbench.access$4(Workbench.java:2494) at org.eclipse.ui.internal.Workbench$7.run(Workbench.java:674) at org.eclipse.core.databinding.observable.Realm.runWithDefault(Realm.java:332) at org.eclipse.ui.internal.Workbench.createAndRunWorkbench(Workbench.java:667) at org.eclipse.ui.PlatformUI.createAndRunWorkbench(PlatformUI.java:149) at org.eclipse.ui.internal.ide.application.IDEApplication.start(IDEApplication.java:123) at org.eclipse.equinox.internal.app.EclipseAppHandle.run(EclipseAppHandle.java:196) at org.eclipse.core.runtime.internal.adaptor.EclipseAppLauncher.runApplication(EclipseAppLauncher.java:110) at org.eclipse.core.runtime.internal.adaptor.EclipseAppLauncher.start(EclipseAppLauncher.java:79) at org.eclipse.core.runtime.adaptor.EclipseStarter.run(EclipseStarter.java:344) at org.eclipse.core.runtime.adaptor.EclipseStarter.run(EclipseStarter.java:179) at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:57) at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43) at java.lang.reflect.Method.invoke(Method.java:616) at org.eclipse.equinox.launcher.Main.invokeFramework(Main.java:622) at org.eclipse.equinox.launcher.Main.basicRun(Main.java:577) at org.eclipse.equinox.launcher.Main.run(Main.java:1410) "VM Thread" prio=10 tid=0x097dd800 nid=0xa9a runnable "GC task thread#0 (ParallelGC)" prio=10 tid=0x08726800 nid=0x1e30 runnable "GC task thread#1 (ParallelGC)" prio=10 tid=0x08727c00 nid=0x1e31 runnable "VM Periodic Task Thread" prio=10 tid=0x097ed400 nid=0xaa0 waiting on condition JNI global references: 2131
上面的stack trace一共可以拆解成5个部分。
第一部分: Full thread dump identifier
2012-08-28 23:16:39 Full thread dump OpenJDK Client VM (19.0-b09 mixed mode, sharing):
可以看到dump时间，虚拟机的信息。
第二部分: Java EE middleware, third party & custom application Threads
"Worker-15" prio=10 tid=0x09b7b400 nid=0x12f2 in Object.wait() [0xb30f9000]
java.lang.Thread.State: TIMED_WAITING (on object monitor)
at java.lang.Object.wait(Native Method)
at org.eclipse.core.internal.jobs.WorkerPool.sleep(WorkerPool.java:188)
- locked <0x6e5f87c0> (a org.eclipse.core.internal.jobs.WorkerPool)
at org.eclipse.core.internal.jobs.WorkerPool.startJob(WorkerPool.java:220)
at org.eclipse.core.internal.jobs.Worker.run(Worker.java:50)
此处是我们最为关心的地方，也是通过stack trace查找问题的地方,每个item的具体含义是
-Thread name: "Worker-15"
-线程优先级： prio=10
-java线程的identifier：tid=0x09b7b400
-native线程的identifier: nid=0x12f2
-线程的状态： in Object.wait() java.lang.Thread.State: TIMED_WAITING (on object monitor)
-线程栈起始地址：[0xb30f9000]
Java线程和Native线程，也就是tid和nid的区别?
VM是加载并执行字节码的机器，这个处理不对应具体硬件，它完全由软件来实现。虚拟机并不是直接跑在宿主的cpu上，中间还隔着宿主的操作系统。
虚拟机的职责是：翻译字节码，把它翻译成一系列的动作（内存操作）和系统调用（比如输入输出）。举个例子：当在java里请求与某台远程机器建立socket连接时，java虚拟机本身是不能完成这一工作的（协议栈是在操作系统内实现的），所以它需要系统调用。要明白这个，还需要知道JRE和JVM的关系。JRE是JVM的环境，一个应用程序就是一个JVM实例。一个机器上一般同时只有一个JRE环境，但是可以同时有多个JVM实例。不同的操作系统和CPU架构有不同的JRE。JRE包含了最基本的类库，当然包括和操作系统打交道的Native库，比如windows下的dll文件，linux下的so文件。JVM实例解析字节码的成JVM中的指令集，再通过Native方法进行操作系统层面的系统调用。下面的图能比较清楚的说明问题。
JVM线程如何创建
当然是通过Native库的 native方法。JVM的线程最终是通过操作系统的线程来完成系统调用的。针对Java开发者来说，发号施令的是JVM 线程，具体干藏活累活的却是操作系统线程。不同操作系统上，JVM创建本地线程依赖的native库也不同。具体可参考下面的的文章。http://jackyhongvip.iteye.com/blog/1338610
JVM线程状态
状态有Runnable、Wait on condition、Waiting for monitor entry 和 in Object.wait()。
Runnable：不解释
Wait on condition:该状态出现在线程等待某个条件的发生。具体是什么原因，可以结合 stacktrace来分析。最常见的情况是线程在等待网络的读写。如果网络数据没准备好，线程就等待在那里。另外一种出现 Wait on condition的常见情况是该线程在 sleep，等待 sleep的时间到了时候，将被唤醒。
Waiting for monitor entry 和 in Object.wait()：这两种情况在多线程的情况下经常出现。Java是通过Monitor来实现线程互斥和协作（有些把Monitor直译成锁，我认为不妥，不是还有Lock嘛）。具体Monitor的深入理解见 http://www.cnblogs.com/tomsheep/archive/2010/06/09/1754419.html
但是可以理解成Monitor是一个对象或者class所拥有的锁，每个对象和class有且仅有一个。见下图.
每个 Monitor在某个时刻，只能被一个线程拥有，该线程就是 “Active Thread”，而其它线程都是 “Waiting Thread”，分别在两个队列 “ Entry Set”和 “Wait Set”里面等候。
在 “ Entry Set”里面的线程都等待拿到Monitor，拿到了线程就成为了Runnable线程，否则就会一直处于处于 “waiting for monitor entry”。一段代码作为例子
public class MyThread implements Runnable{
public void run() {
synchronized(this) {
for (int i = 0; i < 1; i--) {
System.out.println(Thread.currentThread().getName() + " synchronized loop " + i);
}
}
}
public static void main(String[] args) {
MyThread t1 = new MyThread();
Thread ta = new Thread(t1, "A");
Thread tb = new Thread(t1, "B");
ta.start();
tb.start();
}
}
大家一看就知道，B线程肯定是“千年老二“，永远拿不到Monitor了。
对应的stack:
"B" prio=10 tid=0x0969a000 nid=0x11d6 waiting for monitor entry [0x8bb22000] java.lang.Thread.State: BLOCKED (on object monitor) at org.marshal.MyThread.run(MyThread.java:7) - waiting to lock <0x94757078> (a org.marshal.MyThread) at java.lang.Thread.run(Thread.java:636) "A" prio=10 tid=0x09698800 nid=0x11d5 runnable [0x8bb73000] java.lang.Thread.State: RUNNABLE at java.io.FileOutputStream.writeBytes(Native Method) at java.io.FileOutputStream.write(FileOutputStream.java:297) at java.io.BufferedOutputStream.flushBuffer(BufferedOutputStream.java:82) at java.io.BufferedOutputStream.flush(BufferedOutputStream.java:140) - locked <0x947571b0> (a java.io.BufferedOutputStream) at java.io.PrintStream.write(PrintStream.java:449) - locked <0x94757190> (a java.io.PrintStream) at sun.nio.cs.StreamEncoder.writeBytes(StreamEncoder.java:220) at sun.nio.cs.StreamEncoder.implFlushBuffer(StreamEncoder.java:290) at sun.nio.cs.StreamEncoder.flushBuffer(StreamEncoder.java:103) - locked <0x947572a0> (a java.io.OutputStreamWriter) at java.io.OutputStreamWriter.flushBuffer(OutputStreamWriter.java:185) at java.io.PrintStream.write(PrintStream.java:494) - locked <0x94757190> (a java.io.PrintStream) at java.io.PrintStream.print(PrintStream.java:636) at java.io.PrintStream.println(PrintStream.java:773) - locked <0x94757190> (a java.io.PrintStream) at org.marshal.MyThread.run(MyThread.java:8) - locked <0x94757078> (a org.marshal.MyThread) at java.lang.Thread.run(Thread.java:636)
<0x94757078> 就是两个线程争夺的Monitor。
在 “Wait Set”里面的线程都如饥似渴地等待拿到Monitor。他们是怎么进入到“Wait Set”的呢？当一个线程拿到了Monitor，但是在其他资源没有到位的情况下，调用同步锁对象（一般是synchronized()内的对象）的 wait() 方法，放弃了 Monitor，它就进入到了 “Wait Set”队列。只有当其他线程通过notify() 或者 notifyAll()，释放了同步锁后，这个线程才会有机会重新去竞争Monitor。在stack中，它表现的状态是in Object.wait()。修改上面的代码public class WaitThread implements Runnable{
public void run() {
synchronized(this) {
try {
this.wait();
} catch (InterruptedException e) {
// TODO Auto-generated catch block
e.printStackTrace();
}
}
}
public static void main(String[] args) {
WaitThread t1 = new WaitThread();
Thread ta = new Thread(t1, "A");
Thread tb = new Thread(t1, "B");
ta.start();
tb.start();
}
}
对应的stack:
"B" prio=10 tid=0x08173000 nid=0x1304 in Object.wait() [0x8baf2000] java.lang.Thread.State: WAITING (on object monitor) at java.lang.Object.wait(Native Method) - waiting on <0xa9cb50e0> (a org.marshal.WaitThread) at java.lang.Object.wait(Object.java:502) at org.marshal.WaitThread.run(WaitThread.java:8) - locked <0xa9cb50e0> (a org.marshal.WaitThread) at java.lang.Thread.run(Thread.java:636) "A" prio=10 tid=0x08171c00 nid=0x1303 in Object.wait() [0x8bb43000] java.lang.Thread.State: WAITING (on object monitor) at java.lang.Object.wait(Native Method) - waiting on <0xa9cb50e0> (a org.marshal.WaitThread) at java.lang.Object.wait(Object.java:502) at org.marshal.WaitThread.run(WaitThread.java:8) - locked <0xa9cb50e0> (a org.marshal.WaitThread) at java.lang.Thread.run(Thread.java:636)
A和B线程都进入了”wait set“。B线程也拿到过这个Monitor，因为A线程释放过了，这也验证上面的话，他们都在等待得而复失的<0xa9cb50e0>。
基于我们经常讨论到的死锁问题，构造一段代码如下
public class DeadThread implements Runnable{
private Object monitor_A = new Object();
private Object monitor_B = new Object();
public void method_A(){
synchronized(monitor_A) {
synchronized(monitor_B) {
System.out.println(Thread.currentThread().getName()+" invoke method A");
}
}
}
public void method_B(){
synchronized(monitor_B) {
synchronized(monitor_A) {
System.out.println(Thread.currentThread().getName()+" invoke method B");
}
}
}
public void run() {
for(int i=0;i<1;i--){
method_A();
method_B();
}
}
public static void main(String[] args) {
DeadThread t1 = new DeadThread();
Thread ta = new Thread(t1, "A");
Thread tb = new Thread(t1, "B");
ta.start();
tb.start();
}
}
对应的stack:
"B" prio=10 tid=0x0898d000 nid=0x269a waiting for monitor entry [0x8baa2000] java.lang.Thread.State: BLOCKED (on object monitor) at org.marshal.DeadThread.method_A(DeadThread.java:11) - waiting to lock <0xaa4d6f88> (a java.lang.Object) - locked <0xaa4d6f80> (a java.lang.Object) at org.marshal.DeadThread.run(DeadThread.java:28) at java.lang.Thread.run(Thread.java:636) "A" prio=10 tid=0x0898b800 nid=0x2699 waiting for monitor entry [0x8baf3000] java.lang.Thread.State: BLOCKED (on object monitor) at org.marshal.DeadThread.method_B(DeadThread.java:19) - waiting to lock <0xaa4d6f80> (a java.lang.Object) - locked <0xaa4d6f88> (a java.lang.Object) at org.marshal.DeadThread.run(DeadThread.java:29) at java.lang.Thread.run(Thread.java:636)
同时注意到，在stack trace尾部信息
Found one Java-level deadlock: ============================= "B": waiting to lock monitor 0x089615d8 (object 0xaa4d6f88, a java.lang.Object), which is held by "A" "A": waiting to lock monitor 0x08962258 (object 0xaa4d6f80, a java.lang.Object), which is held by "B" Java stack information for the threads listed above: =================================================== "B": at org.marshal.DeadThread.method_A(DeadThread.java:11) - waiting to lock <0xaa4d6f88> (a java.lang.Object) - locked <0xaa4d6f80> (a java.lang.Object) at org.marshal.DeadThread.run(DeadThread.java:28) at java.lang.Thread.run(Thread.java:636) "A": at org.marshal.DeadThread.method_B(DeadThread.java:19) - waiting to lock <0xaa4d6f80> (a java.lang.Object) - locked <0xaa4d6f88> (a java.lang.Object) at org.marshal.DeadThread.run(DeadThread.java:29) at java.lang.Thread.run(Thread.java:636) Found 1 deadlock.
stack中直接报告了Java级别的死锁，够智能吧。
第三部分：HotSpot VM Thread
"Low Memory Detector" daemon prio=10 tid=0x097eb000 nid=0xa9f runnable [0x00000000] java.lang.Thread.State: RUNNABLE "CompilerThread0" daemon prio=10 tid=0x097e9000 nid=0xa9e waiting on condition [0x00000000] java.lang.Thread.State: RUNNABLE
JVM管理的内部线程，这些内部线程都是来执行native操作的，可以不必太关注。
第四部分： HotSpot GC Thread
"GC task thread#0 (ParallelGC)" prio=10 tid=0x08726800 nid=0x1e30 runnable "GC task thread#1 (ParallelGC)" prio=10 tid=0x08727c00 nid=0x1e31 runnable
当虚拟机打开了parallel GC，这些线程就会定期地GC清理。
第五部分：JNI global references count
JNI global references: 2131
JNI (Java Native Interface)全局引用也是受垃圾回收器管理。它的作用是防止回收到那些在native代码上还是被引用的，但是在java代码中却没有存活必要的引用。对于防止因为应用代码中直接使用native库或者第三方的一些监控工具的内存泄漏有非常大的帮助。
参考文献：
http://peiliping.ixiezi.com/2012/04/15/jvm-jstack%E7%94%A8%E6%B3%95/
http://www.cnblogs.com/tomsheep/archive/2010/06/09/1754419.html
http://www.javacodegeeks.com/2012/03/jvm-how-to-analyze-thread-dump.html
http://blog.csdn.net/fenglibing/article/details/6411940