- top获取占用CPU高的进程ID
top
进程pid=26850
top显示信息，需要关注的相关信息：
- load average：反映了任务队列的平均长度。如果此值超过了CPU数量，则表示当前CPU数量不足以处理任务，负载过高
- %us：用户CPU时间百分比。如果此值过高，可能是代码中存在死循环、或是频繁GC等
- %sy：系统CPU时间百分比。如果此值过高，可能是系统线程竞争激烈，上下文切换过多，应当减少线程数
- %wa：等待输入输出CPU时间百分比。如果此值过高，说明系统IO速度过慢，CPU大部分时间都在等待IO完成
- %hi：硬件中断CPU百分比。当硬件中断发生时，CPU会优先去处理，例如网卡收包会产生硬件中断，会优先处理
- swap used：被使用的swap。内存不住时才会用到，此值过高代表系统因为内存不足而导致频繁的换入、换出操作，这样会影响效率，应增大内存量
- %CPU：进程使用CPU的百分比
- top -H -p
获取进程中线程信息
或top 然后H
top -H -p 26850
得到3个高消耗的线程PID:
26857
26853
26855
3.获取最高的线程pid(十进制->十六进制)
点击查看代码
[root@localhost ~]# python -c "print hex(26857)"
0x68e9
或
[root@localhost ~]# printf "0x%x \n" 26857
0x68e9
[root@localhost ~]#
4.查看线程堆栈信息 jstack <进程PID> |grep <线程PID>
可以结合grep参数 -A，-B，-C，控制限制关键字之前、之后、前后 N行
查看对应的堆栈信息，找到可能存在问题的代码
通过jstack，可以查看某个进程的当前线程栈运行情况，定位某个进程的所有线程的当前运行状态、运行代码，以及是否死锁等等
点击查看代码
jstack 26850 |grep 0x68e9
[root@localhost ~]# jstack 26850 |grep -A20 0x68e9
"GC task thread#4 (ParallelGC)" os_prio=0 tid=0x00007fc190025800 nid=0x68e9 runnable
"GC task thread#5 (ParallelGC)" os_prio=0 tid=0x00007fc190027800 nid=0x68ea runnable
"GC task thread#6 (ParallelGC)" os_prio=0 tid=0x00007fc190029800 nid=0x68eb runnable
"GC task thread#7 (ParallelGC)" os_prio=0 tid=0x00007fc19002b000 nid=0x68ec runnable
"VM Periodic Task Thread" os_prio=0 tid=0x00007fc19021c000 nid=0x68f6 waiting on condition
JNI global references: 322
[root@localhost ~]#
- pstack：可以查看某个进程的当前线程栈运行情况
pstack 进程ID - jstat查看gc情况
点击查看代码
[root@localhost ~]# jstat -gc 26850 1000 5
S0C S1C S0U S1U EC EU OC OU MC MU CCSC CCSU YGC YGCT FGC FGCT GCT
903680.0 903680.0 0.0 0.0 904192.0 904192.0 5423104.0 5423017.1 97792.0 92193.5 11520.0 10002.9 1287 85.160 694 7247.812 7332.972
903680.0 903680.0 0.0 0.0 904192.0 904192.0 5423104.0 5423017.1 97792.0 92193.5 11520.0 10002.9 1287 85.160 694 7247.812 7332.972
903680.0 903680.0 0.0 0.0 904192.0 904192.0 5423104.0 5423017.1 97792.0 92193.5 11520.0 10002.9 1287 85.160 694 7247.812 7332.972
903680.0 903680.0 0.0 0.0 904192.0 904192.0 5423104.0 5423017.1 97792.0 92193.5 11520.0 10002.9 1287 85.160 694 7247.812 7332.972
903680.0 903680.0 0.0 0.0 904192.0 821434.1 5423104.0 5423017.1 97792.0 92193.5 11520.0 10002.9 1287 85.160 694 7256.763 7341.923
[root@localhost ~]#
[root@localhost ~]#
[root@localhost ~]# jstat -gcutil 26850 1000 5
S0 S1 E O M CCS YGC YGCT FGC FGCT GCT
0.00 0.00 100.00 100.00 94.28 86.83 1287 85.160 696 7263.066 7348.226
0.00 0.00 100.00 100.00 94.28 86.83 1287 85.160 696 7263.066 7348.226
0.00 0.00 100.00 100.00 94.28 86.83 1287 85.160 696 7263.066 7348.226
0.00 0.00 100.00 100.00 94.28 86.83 1287 85.160 696 7263.066 7348.226
0.00 0.00 100.00 100.00 94.28 86.83 1287 85.160 696 7263.066 7348.226
[root@localhost ~]#
点击查看代码
[root@localhost ~]# jstat -gccause 26850 1s
S0 S1 E O M CCS YGC YGCT FGC FGCT GCT LGCC GCC
0.00 0.00 90.21 99.99 94.31 86.88 1287 85.160 800 8246.108 8331.268 Allocation Failure No GC
0.00 0.00 100.00 99.99 94.31 86.88 1287 85.160 801 8246.108 8331.268 Allocation Failure Ergonomics
0.00 0.00 100.00 99.99 94.31 86.88 1287 85.160 801 8246.108 8331.268 Allocation Failure Ergonomics
0.00 0.00 100.00 99.99 94.31 86.88 1287 85.160 801 8246.108 8331.268 Allocation Failure Ergonomics
0.00 0.00 100.00 99.99 94.31 86.88 1287 85.160 801 8246.108 8331.268 Allocation Failure Ergonomics
0.00 0.00 100.00 99.99 94.31 86.88 1287 85.160 801 8246.108 8331.268 Allocation Failure Ergonomics
0.00 0.00 100.00 99.99 94.31 86.88 1287 85.160 801 8246.108 8331.268 Allocation Failure Ergonomics
0.00 0.00 100.00 99.99 94.31 86.88 1287 85.160 801 8246.108 8331.268 Allocation Failure Ergonomics
0.00 0.00 100.00 99.99 94.31 86.88 1287 85.160 801 8246.108 8331.268 Allocation Failure Ergonomics
0.00 0.00 100.00 99.99 94.33 86.90 1287 85.160 802 8254.123 8339.284 Allocation Failure Ergonomics
0.00 0.00 100.00 99.99 94.33 86.90 1287 85.160 802 8254.123 8339.284 Allocation Failure Ergonomics
0.00 0.00 100.00 99.99 94.33 86.90 1287 85.160 802 8254.123 8339.284 Allocation Failure Ergonomics
0.00 0.00 100.00 99.99 94.33 86.90 1287 85.160 802 8254.123 8339.284 Allocation Failure Ergonomics
0.00 0.00 100.00 99.99 94.33 86.90 1287 85.160 802 8254.123 8339.284 Allocation Failure Ergonomics
^C[root@localhost ~]#
num #instances #bytes class name
1: 110988441 2663722584 com.xxx.demo.proc.ChangeEntry$RowValue
2: 24670473 1012841688 [B
3: 4155237 653276216 [Ljava.lang.Object;
4: 16442770 526168640 java.sql.Timestamp
5: 12352444 296458656 java.lang.Long
6: 8422230 259114432 [C
7: 8418832 202051968 java.lang.String
8: 8241076 197785824 java.lang.Double
9: 4111117 164444680 java.math.BigInteger
10: 4110711 164428440 java.math.BigDecimal
11: 4214003 106784312 [I
12: 4125711 99017064 java.util.ArrayList
通过代码定位，发现：
com.xxx.demo.proc.ChangeEntry这个类对象太多了，大量占用内存。