服务器之家

服务器之家 > 正文

java应用cpu占用过高问题分析及解决方法

时间:2021-01-13 14:45     来源/作者:流子

使用jstack分析java程序cpu占用率过高的问题

1,使用jps查找出java进程的pid,如3707

2,使用top -p 14292 -H观察该进程中所有线程的CPU占用。

?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
[root@cp01-game-dudai-0100.cp01.baidu.com ~]# top -p 14292 -H
top - 22:14:13 up 33 days, 7:29, 4 users, load average: 25.68, 32.11, 33.76
Tasks: 113 total,  2 running, 111 sleeping,  0 stopped,  0 zombie
Cpu(s): 68.3%us, 6.3%sy, 0.0%ni, 20.2%id, 0.1%wa, 0.2%hi, 4.9%si, 0.0%st
Mem: 65965312k total, 65451232k used,  514080k free,  82164k buffers
Swap:  975864k total,  972052k used,   3812k free, 9714400k cached
 PID USER   PR NI VIRT RES SHR S %CPU %MEM  TIME+ COMMAND                                                                                    
15844 root   15  0 6889m 5.7g 4864 S 20.6 9.1 814:13.29 java                                                                                     
15848 root   15  0 6889m 5.7g 4864 S 13.0 9.1 460:25.17 java                                                                                     
15611 root   15  0 6889m 5.7g 4864 S 12.7 9.1 468:17.77 java                                                                                     
15613 root   15  0 6889m 5.7g 4864 S 11.7 9.1 479:40.45 java                                                                                     
15743 root   15  0 6889m 5.7g 4864 S 11.7 9.1 443:04.80 java                                                                                     
15612 root   15  0 6889m 5.7g 4864 S 11.0 9.1 453:43.68 java                                                                                     
15965 root   15  0 6889m 5.7g 4864 S 10.3 9.1 371:00.33 java                                                                                     
15490 root   15  0 6889m 5.7g 4864 S 7.7 9.1 255:32.74 java                                                                                     
15587 root   15  0 6889m 5.7g 4864 S 7.3 9.1 282:27.58 java                                                                                     
15590 root   15  0 6889m 5.7g 4864 S 7.3 9.1 205:48.37 java                                                                                     
15491 root   15  0 6889m 5.7g 4864 R 6.3 9.1 279:09.08 java                                                                                     
15689 root   15  0 6889m 5.7g 4864 S 5.7 9.1 251:42.36 java                                                                                     
16935 root   15  0 6889m 5.7g 4864 S 5.7 9.1 190:34.37 java                                                                                     
15665 root   15  0 6889m 5.7g 4864 S 5.3 9.1 250:07.34 java                                                                                     
16920 root   15  0 6889m 5.7g 4864 S 5.3 9.1 241:34.50 java                                                                                     
15671 root   15  0 6889m 5.7g 4864 S 5.0 9.1 239:49.97 java                                                                                     
15492 root   15  0 6889m 5.7g 4864 S 4.7 9.1 210:23.09 java                                                                                     
14322 root   16  0 6889m 5.7g 4864 S 4.3 9.1 107:39.61 java                                                                                     
14316 root   16  0 6889m 5.7g 4864 S 4.0 9.1 107:18.43 java                                                                                     
14317 root   16  0 6889m 5.7g 4864 S 4.0 9.1 107:29.13 java                                                                                     
15591 root   15  0 6889m 5.7g 4864 S 4.0 9.1 114:34.90 java                                                                                     
14313 root   16  0 6889m 5.7g 4864 S 3.7 9.1 107:12.70 java                                                                                     
14314 root   15  0 6889m 5.7g 4864 S 3.7 9.1 107:28.05 java                                                                                     
14319 root   16  0 6889m 5.7g 4864 S 3.7 9.1 107:27.43 java                                                                                     
14321 root   15  0 6889m 5.7g 4864 S 3.3 9.1 108:01.12 java                                                                                     
15589 root   15  0 6889m 5.7g 4864 R 3.0 9.1 109:01.91 java                                                                                     
15615 root   15  0 6889m 5.7g 4864 S 3.0 9.1 114:55.29 java                                                                                     
16808 root   15  0 6889m 5.7g 4864 S 2.7 9.1 279:05.03 java                                                                                     
14315 root   15  0 6889m 5.7g 4864 S 2.0 9.1 107:45.00 java                                                                                     
14320 root   15  0 6889m 5.7g 4864 S 2.0 9.1 107:48.30 java                                                                                     
15489 root   15  0 6889m 5.7g 4864 S 1.7 9.1 57:38.46 java                                                                                     
15670 root   15  0 6889m 5.7g 4864 S 1.3 9.1  5:55.43 java                                                                                     
14318 root   15  0 6889m 5.7g 4864 S 0.7 9.1 107:45.88 java                                                                                     
14826 root   15  0 6889m 5.7g 4864 S 0.7 9.1 25:07.64 java

 

3,找出CPU消耗较多的线程id,如15844,将15844转换为16进制0x3de4,注意是小写哦

4,使用jstack 14292|grep -A 10 0x3de4来查询出具体的线程状态。

?
1
2
3
4
5
6
7
8
9
10
11
[root@cp01-game-dudai-0100.cp01.baidu.com ~]# jstack 14292|grep -A 10 0x3de4
"pool-52-thread-1" prio=10 tid=0x000000005a08e000 nid=0x3de4 waiting on condition [0x00002ae63d917000]
  java.lang.Thread.State: WAITING (parking)
    at sun.misc.Unsafe.park(Native Method)
    - parking to wait for <0x00000006f9a0a110> (a java.util.concurrent.locks.AbstractQueuedSynchronizer$ConditionObject)
    at java.util.concurrent.locks.LockSupport.park(LockSupport.java:156)
    at java.util.concurrent.locks.AbstractQueuedSynchronizer$ConditionObject.await(AbstractQueuedSynchronizer.java:1987)
    at java.util.concurrent.LinkedBlockingQueue.take(LinkedBlockingQueue.java:399)
    at java.util.concurrent.ThreadPoolExecutor.getTask(ThreadPoolExecutor.java:947)
    at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:907)
    at java.lang.Thread.run(Thread.java:662)

通过这些线程状态便可基本定位问题之所在。

解决办法:

方法1

1.jps 获取Java进程的PID。

2.jstack pid >> java.txt 导出CPU占用高进程的线程栈。

3.top -H -p PID 查看对应进程的哪个线程占用CPU过高。

4.echo “obase=16; PID” | bc 将线程的PID转换为16进制,大写转换为小写。

5.在第二步导出的Java.txt中查找转换成为16进制的线程PID。找到对应的线程栈。

6.分析负载高的线程栈都是什么业务操作。优化程序并处理问题。

方法2

1.使用top 定位到占用CPU高的进程PID

top

通过ps aux | grep PID命令

2.获取线程信息,并找到占用CPU高的线程

ps -mp pid -o THREAD,tid,time | sort -rn

3.将需要的线程ID转换为16进制格式

printf "%x\n" tid

4.打印线程的堆栈信息

jstack pid |grep tid -A 30

总结

以上就是本文关于java应用cpu占用过高问题分析及解决方法的全部内容,希望对大家有所帮助,如果有什么疑问,可以随时留言,小编会及时回复大家的。感谢朋友们对服务器之家网站的支持。

原文链接:http://blog.csdn.net/jiangguilong2000/article/details/17971247

标签:

相关文章

热门资讯

2020微信伤感网名听哭了 让对方看到心疼的伤感网名大全
2020微信伤感网名听哭了 让对方看到心疼的伤感网名大全 2019-12-26
Intellij idea2020永久破解,亲测可用!!!
Intellij idea2020永久破解,亲测可用!!! 2020-07-29
背刺什么意思 网络词语背刺是什么梗
背刺什么意思 网络词语背刺是什么梗 2020-05-22
歪歪漫画vip账号共享2020_yy漫画免费账号密码共享
歪歪漫画vip账号共享2020_yy漫画免费账号密码共享 2020-04-07
苹果12mini价格表官网报价 iPhone12mini全版本价格汇总
苹果12mini价格表官网报价 iPhone12mini全版本价格汇总 2020-11-13
返回顶部