Linux重复执行命令之watch命令10个案例

1. Linux watch命令

Linux日常维护中，经常需要重复执行某些指令。比如，重复执行ip addr检查dhcp和网络是否正常，是否可以正常获取IP地址。这些watch命令相当有用，同时它也支持定时重复执行某命令。默认为2秒重复执行一次。

2. 每隔1秒查看网络电缆是否已经连接好

[root@zcwyou ~]# watch -n 1 "ethtool ens160 |tail -n 1"

3. 每隔1秒查看TCP连接数，怀疑受到网络攻击时非常实用

[root@zcwyou ~]# watch -n 1 "netstat -ntp |wc -l"

4. 每隔30秒查看系统负载

[root@zcwyou ~]# watch -n 30 'cat /proc/loadavg'

5. 每隔1秒查看http连接数，怀疑受到CC攻击时非常实用

[root@zcwyou ~]# watch -n 1 -d 'pstree|grep http'

6. 实时跟踪系统日志的变化

[root@zcwyou ~]# watch tail /var/log/messages

Every 2.0s: tail /var/log/messages Wed Nov 14 09:54:51 2018

Nov 14 09:54:36 controller NetworkManager[764]: [1542160476.7311] device (p1p1): state change: ip-config -> failed (reason ‘i
p-config-unavailable’, sys-iface-state: ‘managed’)
Nov 14 09:54:36 controller NetworkManager[764]: [1542160476.7314] manager: NetworkManager state is now CONNECTED_LOCAL
Nov 14 09:54:36 controller NetworkManager[764]: [1542160476.7317] device (p1p1): Activation: failed for connection ‘p1p1’
Nov 14 09:54:36 controller NetworkManager[764]: [1542160476.7322] device (p1p1): state change: failed -> disconnected (reason
‘none’, sys-iface-state: ‘managed’)

7. 跟踪系统运行时间和负载

[root@centos7 ~]# watch uptime

Every 2.0s: uptime Wed Nov 14 09:53:53 2018

09:53:53 up 9 days, 17:16, 1 user, load average: 0.80, 0.73, 0.65

8. 每隔3秒输出内存使用情况

选项-d会对比上一次结果，不同的地方会以高亮显示

[root@centos7 ~]# watch -n 3 -d free -m

输出结果如下：

Every 3.0s: free -m Wed Nov 14 09:53:15 2018

          total        used        free  shared  buff/cache   available

Mem: 31996 19618 254 95 12124 11824
Swap: 4095 111 3984

9. 每隔20秒输出系统负载情况，并高亮显示差异点

[root@zcwyou ~]# watch -n 20 -d uptime

Every 20.0s: uptime Wed Nov 14 09:55:36 2018
09:55:36 up 9 days, 17:18, 1 user, load average: 0.62, 0.69, 0.65

10. 查看watch的版本号。

[root@zcwyou ~]# watch -h [--help]