主题ps查看和调整优先级
一查看优先级
nice值越大优先级越低
[root@centos72 ~]# ps axo pid,cmd,ni | head
PID CMD NI
1 /usr/lib/systemd/systemd -- 0
2 [kthreadd] 0
3 [ksoftirqd/0] 0
5 [kworker/0:0H] -20
7 [migration/0] -
8 [rcu_bh] 0
9 [rcu_sched] 0
10 [lru-add-drain] -20
11 [watchdog/0] -
[root@centos72 ~]# ps axo pid,cmd,ni | tail
8406 [kworker/1:1] 0
8457 sshd: root@pts/3 0
8465 -bash 0
8485 dd if=/dev/zero of=/dev/nul 0
8573 [kworker/1:1H] -20
10541 [kworker/0:2] 0
11025 [kworker/0:1] 0
11295 [kworker/0:0] 0
11297 ps axo pid,cmd,ni 0
11298 tail 0
开启一个进程
[root@centos72 ~]# ping -f 127.0.0.1
PING 127.0.0.1 (127.0.0.1) 56(84) bytes of data.
^C
--- 127.0.0.1 ping statistics ---
4860967 packets transmitted, 4860967 received, 0% packet loss, time 44792ms
rtt min/avg/max/mdev = 0.001/0.003/14.741/0.028 ms, pipe 2, ipg/ewma 0.009/0.009 ms
按照CPU进行排序，CPU占用过高
[root@centos72 ~]# ps axo pid,cmd,ni,%cpu k %cpu | tail
1142 -bash 0 0.0
1230 [kworker/0:1] 0 0.0
1253 [kworker/0:2] 0 0.0
1255 sshd: root@pts/1 0 0.0
1259 -bash 0 0.0
1338 ps axo pid,cmd,ni,%cpu k %c 0 0.0
1339 tail 0 0.0
1294 [kworker/0:0] 0 0.1
906 sshd: root@pts/0 0 0.2
1283 ping -f 127.0.0.1 0 95.9
二调整优先级
注意此时进程正在运行
[root@centos72 ~]# ps axo pid,cmd,ni,%cpu k %cpu | grep ping
1355 grep --color=auto ping 0 0.0
1283 ping -f 127.0.0.1 0 96.0
[root@centos72 ~]# renice -n -20 1283
1283 (process ID) old priority 0, new priority -20
[root@centos72 ~]# ps axo pid,cmd,ni,%cpu k %cpu | grep ping
1358 grep --color=auto ping 0 0.0
1283 ping -f 127.0.0.1 -20 95.8
注意此时进程正在运行
[root@centos72 ~]# ping -f 127.0.0.1
PING 127.0.0.1 (127.0.0.1) 56(84) bytes of data.
.^
--- 127.0.0.1 ping statistics ---
7827881 packets transmitted, 7827881 received, 0% packet loss, time 58781ms
rtt min/avg/max/mdev = 0.001/0.002/13.661/0.011 ms, pipe 2, ipg/ewma 0.007/0.006 ms
在执行命令的时候直接调整优先级
使用nice命令指定
[root@centos72 ~]# nice -n -5 ping -f 127.0.0.1
PING 127.0.0.1 (127.0.0.1) 56(84) bytes of data.
^C
--- 127.0.0.1 ping statistics ---
3272228 packets transmitted, 3272228 received, 0% packet loss, time 26347ms
rtt min/avg/max/mdev = 0.001/0.003/8.853/0.015 ms, pipe 2, ipg/ewma 0.008/0.005 ms
[root@centos72 ~]#
[root@centos72 ~]# ps axo pid,cmd,ni,%cpu | grep ping
1158 ping -f 127.0.0.1 -5 94.6
1161 grep --color=auto ping 0 0.0
[root@centos72 ~]# nice -n 5 ping -f 127.0.0.1
PING 127.0.0.1 (127.0.0.1) 56(84) bytes of data.
[root@centos72 ~]# ps axo pid,cmd,ni,%cpu | grep ping
1162 ping -f 127.0.0.1 5 110
1164 grep --color=auto ping 0 0.0
优先级高并不一定更快
[root@centos72 ~]# nice -n -11 ping 127.0.0.1
PING 127.0.0.1 (127.0.0.1) 56(84) bytes of data.
64 bytes from 127.0.0.1: icmp_seq=1 ttl=64 time=0.019 ms
64 bytes from 127.0.0.1: icmp_seq=2 ttl=64 time=0.030 ms
64 bytes from 127.0.0.1: icmp_seq=3 ttl=64 time=0.026 ms
^C
--- 127.0.0.1 ping statistics ---
3 packets transmitted, 3 received, 0% packet loss, time 1999ms
rtt min/avg/max/mdev = 0.019/0.025/0.030/0.004 ms
[root@centos72 ~]# ping 127.0.0.1
PING 127.0.0.1 (127.0.0.1) 56(84) bytes of data.
64 bytes from 127.0.0.1: icmp_seq=1 ttl=64 time=0.028 ms
64 bytes from 127.0.0.1: icmp_seq=2 ttl=64 time=0.027 ms
64 bytes from 127.0.0.1: icmp_seq=3 ttl=64 time=0.022 ms
64 bytes from 127.0.0.1: icmp_seq=4 ttl=64 time=0.023 ms
64 bytes from 127.0.0.1: icmp_seq=5 ttl=64 time=0.023 ms
^C
--- 127.0.0.1 ping statistics ---
5 packets transmitted, 5 received, 0% packet loss, time 3999ms
rtt min/avg/max/mdev = 0.022/0.024/0.028/0.006 ms
三注意事项——优先级之间的关系
（一）ni: nice值的优先级和pri优先级是相反的
nice值的优先级越高，那么对应的pri优先级越低
[root@centos72 ~]# ps axo pid,cmd,ni,%cpu,pri | grep ping
1283 ping -f 127.0.0.1 -20 96.2 39
1360 grep --color=auto ping 0 0.0 19
[root@centos72 ~]# ps axo pid,cmd,ni,%cpu,pri | tail
1099 [kworker/0:1] 0 0.0 19
1101 sshd: root@pts/2 0 0.7 19
1105 -bash 0 0.0 19
1143 [kworker/1:1] 0 0.0 19
1159 [kworker/1:2] 0 0.0 19
1165 [kworker/0:0] 0 0.0 19
1183 [kworker/0:2] 0 0.0 19
1184 [kworker/1:0] 0 0.0 19
1190 ps axo pid,cmd,ni,%cpu,pri 0 0.0 19
1191 -bash 0 0.0 19
[root@centos72 ~]# nice -n 10 ping 127.0.0.1
PING 127.0.0.1 (127.0.0.1) 56(84) bytes of data.
64 bytes from 127.0.0.1: icmp_seq=1 ttl=64 time=0.025 ms
64 bytes from 127.0.0.1: icmp_seq=2 ttl=64 time=0.023 ms
64 bytes from 127.0.0.1: icmp_seq=3 ttl=64 time=0.027 ms
64 bytes from 127.0.0.1: icmp_seq=4 ttl=64 time=0.026 ms
64 bytes from 127.0.0.1: icmp_seq=5 ttl=64 time=0.026 ms
64 bytes from 127.0.0.1: icmp_seq=6 ttl=64 time=0.027 ms
64 bytes from 127.0.0.1: icmp_seq=7 ttl=64 time=0.026 ms
64 bytes from 127.0.0.1: icmp_seq=8 ttl=64 time=0.057 ms
64 bytes from 127.0.0.1: icmp_seq=9 ttl=64 time=0.023 ms
64 bytes from 127.0.0.1: icmp_seq=10 ttl=64 time=0.025 ms
64 bytes from 127.0.0.1: icmp_seq=11 ttl=64 time=0.024 ms
64 bytes from 127.0.0.1: icmp_seq=12 ttl=64 time=0.026 ms
64 bytes from 127.0.0.1: icmp_seq=13 ttl=64 time=0.024 ms
64 bytes from 127.0.0.1: icmp_seq=14 ttl=64 time=0.037 ms
64 bytes from 127.0.0.1: icmp_seq=15 ttl=64 time=0.023 ms
^C
--- 127.0.0.1 ping statistics ---
15 packets transmitted, 15 received, 0% packet loss, time 14000ms
rtt min/avg/max/mdev = 0.023/0.027/0.057/0.011 ms
[root@centos72 ~]# ps axo pid,cmd,ni,%cpu,pri | tail
1101 sshd: root@pts/2 0 0.7 19
1105 -bash 0 0.0 19
1143 [kworker/1:1] 0 0.0 19
1159 [kworker/1:2] 0 0.0 19
1165 [kworker/0:0] 0 0.0 19
1183 [kworker/0:2] 0 0.0 19
1184 [kworker/1:0] 0 0.0 19
1192 ping 127.0.0.1 10 0.0 9
1193 ps axo pid,cmd,ni,%cpu,pri 0 0.0 19
1194 tail 0 0.0 19
[root@centos72 ~]# nice -n -10 ping 127.0.0.1
PING 127.0.0.1 (127.0.0.1) 56(84) bytes of data.
64 bytes from 127.0.0.1: icmp_seq=1 ttl=64 time=0.021 ms
64 bytes from 127.0.0.1: icmp_seq=2 ttl=64 time=0.028 ms
64 bytes from 127.0.0.1: icmp_seq=3 ttl=64 time=0.027 ms
64 bytes from 127.0.0.1: icmp_seq=4 ttl=64 time=0.023 ms
64 bytes from 127.0.0.1: icmp_seq=5 ttl=64 time=0.023 ms
64 bytes from 127.0.0.1: icmp_seq=6 ttl=64 time=0.023 ms
^C
--- 127.0.0.1 ping statistics ---
6 packets transmitted, 6 received, 0% packet loss, time 4999ms
rtt min/avg/max/mdev = 0.021/0.024/0.028/0.003 ms
[root@centos72 ~]# ps axo pid,cmd,ni,%cpu,pri | tail
1101 sshd: root@pts/2 0 0.6 19
1105 -bash 0 0.0 19
1143 [kworker/1:1] 0 0.0 19
1159 [kworker/1:2] 0 0.0 19
1165 [kworker/0:0] 0 0.0 19
1183 [kworker/0:2] 0 0.0 19
1184 [kworker/1:0] 0 0.0 19
1195 ping 127.0.0.1 -10 0.0 29
1196 ps axo pid,cmd,ni,%cpu,pri 0 0.0 19
1197 tail 0 0.0 19
注意帮助文档的此说明并不准确
[root@centos72 ~]# man ps | grep PRI
pri PRI priority of the process. Higher number means lower priority.
rtprio RTPRIO realtime priority.
数字越大优先级越高
所以帮助文档是有一定的bug的，不要迷信权威
[root@centos72 ~]# ps -C ping o pid,cmd,ni,%cpu,pri
PID CMD NI %CPU PRI
1246 ping 127.0.0.1 10 0.0 9
1248 ping 127.0.0.1 -10 0.0 29
在6上没有PRI的相关说明
[root@centos65 ~]# man ps | grep PRI
rtprio RTPRIO realtime priority.
（二）pri的优先级的值和早期系统优先级相反
也就是从139到0，早期的系统优先级是从0到139
所以pri的值越大优先级越高
[root@centos72 ~]# nice -n 30 ping 127.0.0.1
[root@centos72 ~]# ps axo pid,cmd,ni,%cpu,pri | tail
1058 ping 8.8.8.8 0 0.0 19
1059 sshd: root@pts/1 0 0.1 19
1063 -bash 0 0.0 19
1086 sshd: root@pts/2 0 0.6 19
1090 -bash 0 0.0 19
1110 [kworker/1:1] 0 0.0 19
1113 [kworker/0:0] 0 0.0 19
1114 ping 127.0.0.1 19 0.0 0
1118 ps axo pid,cmd,ni,%cpu,pri 0 0.0 19
1119 tail 0 0.0 19
（三）实时优先级rtprio和nice优先级关系
从下图可以看出实时优先级和nice优先级对应系统优先级的一部分，实时优先级和nice优先级是没有交集的
ncie优先级又被称为动态优先级，比如占用的内存太多了就会自动的把优先级调小。
实时优先级是静态的优先级，指定好了就固定不变了。
调整实时优先级使用chrt
rtprio: 实时优先级
因为nice优先级不对应任何的实时优先级，所以无法显示实时优先级
[root@centos72 ~]# nice -n -20 ping 127.0.0.1
PING 127.0.0.1 (127.0.0.1) 56(84) bytes of data.
64 bytes from 127.0.0.1: icmp_seq=1 ttl=64 time=0.034 ms
64 bytes from 127.0.0.1: icmp_seq=2 ttl=64 time=0.026 ms
64 bytes from 127.0.0.1: icmp_seq=3 ttl=64 time=0.022 ms
^C
--- 127.0.0.1 ping statistics ---
3 packets transmitted, 3 received, 0% packet loss, time 2000ms
rtt min/avg/max/mdev = 0.022/0.027/0.034/0.006 ms
[root@centos72 ~]# ps -C ping o pid,cmd,ni,%cpu,pri,rtprio | tail
PID CMD NI %CPU PRI RTPRIO
1260 ping 127.0.0.1 -20 0.0 39 -
7 [migration/0] - 0.0 139 99中139是最高的优先级，值越大优先级越高，
因为是pri优先级，和系统优先级是相反的
[root@centos72 ~]# ps axo pid,cmd,ni,%cpu,pri,rtprio | head
PID CMD NI %CPU PRI RTPRIO
1 /usr/lib/systemd/systemd -- 0 0.0 19 -
2 [kthreadd] 0 0.0 19 -
3 [ksoftirqd/0] 0 0.0 19 -
5 [kworker/0:0H] -20 0.0 39 -
6 [kworker/u256:0] 0 0.0 19 -
7 [migration/0] - 0.0 139 99
8 [rcu_bh] 0 0.0 19 -
9 [rcu_sched] 0 0.0 19 -
10 [lru-add-drain] -20 0.0 39 -
看门狗监控内核状态的，最高优先级
就像国家的安全部分监控其他部门，比如明朝的锦衣卫
[root@centos72 ~]# ps axo pid,cmd,ni,%cpu,pri,rtprio | grep watch
11 [watchdog/0] - 0.0 139 99
12 [watchdog/1] - 0.0 139 99
1272 grep --color=auto watch 0 0.0 19 -