linux（unix）下java应用占用cpu过高查找方法

先说说linux下怎么处理

1、使用top命令查看CPU使用情况

top - 17:07:58 up 2 days,  7:36,  4 users,  load average: 1.49, 0.45, 0.20
Tasks: 115 total,   2 running, 105 sleeping,   8 stopped,   0 zombie
Cpu(s):  83.7%us,  2.0%sy,  0.0%ni, 14.4%id,  0.0%wa,  0.0%hi,  0.0%si,  0.0%st
Mem:   1035160k total,   976428k used,    58732k free,    43772k buffers
Swap:  2104472k total,  1138312k used,   966160k free,   289436k cached

  PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND                                                                                       
2902 test       15   0  403m 146m 9160 S  83.0 14.5   3:17.50 java                                                                                           
2386 root      17   0  1860  480  464 S  0.3  0.0   0:46.27 hald-addon-stor  

2、可以看出，这是一个java进程，属于test用户，使用ps命令可以查看该进程的具体信息

ps -ef|grep 2902

test@linux-nc6d:~> ps -ef|grep 2902
test       2902     1  1 12:26 ?        00:03:20 java -Test -Xms256m -Xmx256m -jar ./plugins/org.eclipse.equinox.launcher_1.3.0.v20130327-1440.jar -clean -refresh
test      17340  9586  0 17:12 pts/6    00:00:00 grep 2902

3、现在已经确定了是2902这个进程占用了大量CPU，继续查看该问题是由2902进程中的哪个线程导致

top -Hp 2902

top - 17:20:19 up 2 days,  7:48,  4 users,  load average: 0.38, 0.66, 0.47
Tasks:  46 total,   0 running,  46 sleeping,   0 stopped,   0 zombie
Cpu(s):  82.0%us,  1.7%sy,  0.0%ni, 15.7%id,  0.3%wa,  0.3%hi,  0.0%si,  0.0%st
Mem:   1035160k total,   941700k used,    93460k free,    65152k buffers
Swap:  2104472k total,  1149404k used,   955068k free,   244544k cached

  PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND                                                                                       
17797 test       15   0  403m 146m 9160 S  83.0 14.5   0:02.67 java                                                                                           
2906 test       16   0  403m 146m 9160 S  0.3 14.5   0:06.96 java                                                                                           
2902 test       15   0  403m 146m 9160 S  0.0 14.5   0:00.00 java                                                                                           
2903 test       16   0  403m 146m 9160 S  0.0 14.5   0:01.09 java                                                                                           
2907 test       16   0  403m 146m 9160 S  0.0 14.5   0:00.60 java                                                                                           
2908 test       16   0  403m 146m 9160 S  0.0 14.5   0:00.53 java  

4、可以看到17797线程占用了几乎全部的CPU资源， 17797在干什么，让打开堆栈一探究竟
使用jstack拿到进程的执行情况，再把十进制的17797换算成16进制，即4585（window计算器调整为程序员模式，自由换算）
5、现在，从我们拿到的2902.txt中搜索关键字“4585”，线程堆栈信息到手~

jstack -l 2902 > 2902.txt

"Worker of StableTimer" daemon prio=10 tid=0x9f801000 nid=0x4585 waiting on condition [0x9fece000]
   java.lang.Thread.State: TIMED_WAITING (parking)
        at sun.misc.Unsafe.park(Native Method)
        - parking to wait for  <0xa6b51090> (a java.util.concurrent.Semaphore$NonfairSync)
        at java.util.concurrent.locks.LockSupport.parkNanos(LockSupport.java:226)
        at java.util.concurrent.locks.AbstractQueuedSynchronizer.doAcquireSharedNanos(AbstractQueuedSynchronizer.java:1033)
        at java.util.concurrent.locks.AbstractQueuedSynchronizer.tryAcquireSharedNanos(AbstractQueuedSynchronizer.java:1326)
        at java.util.concurrent.Semaphore.tryAcquire(Semaphore.java:588)
        at test.StableTimer$2.run(StableTimer.java:87)
        at java.lang.Thread.run(Thread.java:745)

6、根据堆栈信息，去应用中查找问题吧，没准是个死循环~

另外如果拿到的线程信息是与GC相关的，那么恭喜，有可能该问题由内存泄露导致的，淡定的去分析内存对象吧。

下面继续谈谈AIX下出现这种问题时该怎么办

解决思路基本一致

1、使用topas命令查看cpu占用很高

    Name            PID  CPU%  PgSp Owner    
    java        6029454  75.1 172.2 abs   
    topas       5505138   0.0   4.4 abs   
    topas      13500534   0.0   4.5 abs

2、此时使用kill -3 6029454 会在abs目录下生成javacore.20131213.085135.8978506.0004.txt文件

3、使用ps -mp 6029454 -o THREAD 查看abs进程中占用cpu最高的线程（CP列（表示CPU占用率）），如下的18612279 25755729

USER      PID     PPID       TID S  CP PRI SC    WCHAN        F     TT BND COMMAND
         abs  8978506        1         - A 442  60 91        *   242001      -  -1 java -Dabs_node=CS1 -Xms1024m -Xmx1024m -verbose:gc -Xverbosegclog:/abslog/fjnxcslog1/gclog/20131212171643.gc -XX:+PrintGCDetails -XX:+PrintGCDateStamps -XX:+PrintGCApplicationStoppedTime -XX:+PrintHeapAtGC -XX:+PrintHeapAtGCExtended -XX:+HeapDumpOnOutOfMemoryError -XX:PermSize=512m -XX:MaxPermSize=512m -Xdump:heap:events=user -jar ./plugins/org.eclipse.equinox.launcher_1.2.0.v20110502.jar -clean -refresh
           -        -        -  13172911 S   0  82  1 f1000f0a1000c940  8410400      -  -1 -
           -        -        -  17760291 R   2  83  0        -   400000      -  -1 -
           -        -        -  18612279 S  57 120  1 f1000f0a10011c40  8410400      -  -1 -
           -        -        -  24772643 S   0  78  1 f1000f0a10017a40  8410400      -  -1 -
           -        -        -  25755729 S  57  60  1 f1000f0a10018940  8410400      -  -1 -

4、使用Integer.toHexString(18612279)把执行线程转换为16进制 11c0037

5、在javacore文件中查找11c0037，即可搜索到对应堆栈信息，可以看出本次问题出现在T671013.java:79行。找到T671013.java后即可查找到占用原因

3XMTHREADINFO      "Acceptor0.Processor2.Worker16" J9VMThread:0x00000100163C4300, j9thread_t:0x00000100135046E0, java/lang/Thread:0x07000000014C34D0, state:CW, prio=5
        3XMJAVALTHREAD            (java/lang/Thread getId:0x1F5, isDaemon:false)
        3XMTHREADINFO1            (native thread ID:0x11C0037, native priority:0x5, native policy:UNKNOWN)
        3XMTHREADINFO3           Java callstack:
        4XESTACKTRACE                at trade/zhqy/t671013/T671013.tradeInit(T671013.java:79(Compiled Code))
        4XESTACKTRACE                at lib/controller/tradeHandler/TransactHandler.tradeInit(TransactHandler.java:33)
        4XESTACKTRACE                at lib/controller/common/Common.tradeInit(Common.java:21)