使用`ps`命令，当处理器有多核时，`%cpu`列是什么意思？

Question

请看我的CPU架构信息。

root@jai [~]# lscpu
Architecture:          x86_64
CPU op-mode(s):        32-bit, 64-bit
Byte Order:            Little Endian
CPU(s):                16
On-line CPU(s) list:   0-15
Thread(s) per core:    1
Core(s) per socket:    1
Socket(s):             16
NUMA node(s):          1
Vendor ID:             GenuineIntel
CPU family:            6
Model:                 86
Model name:            Intel(R) Xeon(R) CPU D-1541 @ 2.10GHz
Stepping:              3
CPU MHz:               2099.998
BogoMIPS:              4199.99
Hypervisor vendor:     KVM
Virtualization type:   full
L1d cache:             32K
L1i cache:             32K
L2 cache:              4096K
L3 cache:              16384K
NUMA node0 CPU(s):     0-15

我的网站性能很差，有时工作时断时续，一次不能完全加载页面，但几秒钟后就能加载所有页面。完全是间歇性的，而且在不同地点的不同电脑和手机上也发生了同样的情况。我怀疑这个问题是由于超过了我的CPU容量，无法处理我当时运行的流量和处理。于是我在遇到问题的时候运行了以下命令。看到结果。

root@cup [~]# ps -eo pid,ppid,cmd,%mem,%cpu --sort=-%cpu | more
  PID  PPID CMD                         %MEM %CPU
26468  1499 php-fpm: pool m_jaimemontoya_n  1.0 57.3
26463  1499 php-fpm: pool m_jaimemontoya_n  0.7 42.3
26553  1499 php-fpm: pool m_jaimemontoya_n  0.6 38.5
26502  1499 php-fpm: pool m_jaimemontoya_n  0.2 35.6
26190  1499 php-fpm: pool m_jaimemontoya_n  0.3 29.4
18242  1499 php-fpm: pool m_jaimemontoya_n  1.1 22.8
19045  1499 php-fpm: pool m_jaimemontoya_n  1.0 20.6
18437  1499 php-fpm: pool m_jaimemontoya_n  0.6 20.2
18269  1499 php-fpm: pool m_jaimemontoya_n  1.1 18.5
18289  1499 php-fpm: pool m_jaimemontoya_n  0.9 13.1
19042  1499 php-fpm: pool m_jaimemontoya_n  1.1 11.9
26906     1 /usr/sbin/exim -Mc 1jTsTl-0  0.0  7.0
 8546  8014 /usr/sbin/mysqld --basedir=  1.1  6.7
26872     1 /usr/sbin/exim -Mc 1jTsTl-0  0.0  6.0
26877     1 /usr/sbin/exim -Mc 1jTsTl-0  0.0  6.0
26869     1 /usr/sbin/exim -Mc 1jTsTl-0  0.0  5.0
26875 21851 /usr/sbin/exim -qG           0.0  5.0
26885     1 /usr/sbin/exim -Mc 1jTsTl-0  0.0  5.0
26891     1 /usr/sbin/exim -Mc 1jTsTl-0  0.0  5.0
26895     1 /usr/sbin/exim -Mc 1jTsTl-0  0.0  5.0
26888     1 /usr/sbin/exim -Mc 1jTsTl-0  0.0  4.0
26903     1 /usr/sbin/exim -Mc 1jTsTl-0  0.0  4.0
26880 26875 /usr/sbin/exim -qG           0.0  3.0
26881 26869 /usr/sbin/exim -Mc 1jTsTl-0  0.0  3.0
26882 26872 /usr/sbin/exim -Mc 1jTsTl-0  0.0  3.0
26899 26888 /usr/sbin/exim -Mc 1jTsTl-0  0.0  3.0
17419  1499 php-fpm: pool jaimemontoya_com  0.0  2.5
26814     1 /usr/sbin/exim -Mc 1jTsTj-0  0.0  2.5
26849     1 /usr/sbin/exim -Mc 1jTsTk-0  0.0  2.5
21290  1499 php-fpm: pool jaimemontoya_com  0.0  2.1
14959  1499 php-fpm: pool jaimemontoya_com  0.0  2.0
16122  1499 php-fpm: pool jaimemontoya_com  0.0  2.0
17085  1499 php-fpm: pool jaimemontoya_com  0.0  2.0
22367  1499 php-fpm: pool jaimemontoya_com  0.0  2.0
26826     1 /usr/sbin/exim -Mc 1jTsTk-0  0.0  2.0
26859 26849 /usr/sbin/exim -Mc 1jTsTk-0  0.0  2.0
26884 26877 /usr/sbin/exim -Mc 1jTsTl-0  0.0  2.0
26902 26895 /usr/sbin/exim -Mc 1jTsTl-0  0.0  2.0
18723  1499 php-fpm: pool jaimemontoya_com  0.0  1.7
21456  1499 php-fpm: pool jaimemontoya_com  0.0  1.7
21975  1499 php-fpm: pool jaimemontoya_com  0.0  1.7
13578  1499 php-fpm: pool jaimemontoya_com  0.0  1.6

我想知道如何解释 %CPU 列。如果我把下面的值加起来，我得到404.2的总和。

57.3+42.3+38.5+35.6+29.4+22.8+20.6+20.2+18.5+13.1+11.9+7.0+6.7+6.0+5.0+5.0+5.0+5.0+5.0+4.0+4.0+3.0+3.0+3.0+2.5+2.5+2.5+2.1+2.0+2.0+2.0+2.0+2.0+2.0+2.0+2.0+1.7+1.7+1.7+1.6

考虑到我有16个CPU，我怎么能把它解释为： 404.2 中的所有值相加后的结果。%CPU 所有进程的列？我有一台服务器，内存30GB，CPU为8核16线程。谢谢你，我的服务器有30GB内存，8核16线程CPU。

注：过去我的服务器日志上经常收到这个警告。WARNING: [pool m_jaimemontoya_com] server reached max_children setting (5), consider raising it 所以我把它提高到32。出于这个原因，你会看到超过5个处理的 php-fpm 同时运行。

更新1:

使用 man lscpu，我看到了这个。

COLUMNS
       CPU    The logical CPU number of a CPU as used by the Linux kernel.

Answer 1

我要求AbraCadaver把他的评论写成答案，因为那是我的解决方案。他没有这样做。我把他的评论抄在下面，作为我问题的答案。

"这是个糟糕的数据点 ps CPU%是指CPU的使用率，目前表示为一个进程在整个生命周期中运行时间的百分比。可以看一下 top 或其他效用。- AbraCadaver Apr 29 at 20:41"