周末写了个 monitor filesystem shell ( crontab 定时监控)

现在公司的 oracle 文件的 归档filesystem 下太大了，而且要是录用达到100% oracle 的用户是connect 不上去的，于是决定 写个监控的shell:

 

思想： 归档的日志有大有小，当filesystem 在20% 一下就可以认为正常，不要删除，否则就要delete old files.

 

1:好下面来看以下shell: (是放在/tmp/test/monitor.sh)

 

#! /bin/ksh

 

# Function: This shell is to monitor the filesystem.

# Date: Dev 05 2010 sunny Sunday

 

# step1 To to check the filesytem, if it more than 20%，then delete the old files, otherwise to do nothing.

 

df -g > checkfile

grep '/dev/arch' checkfile > ch01

awk '{$print $4}'  ch01 > ch02

ch03=ch02

ch04 = `cat $ch03`

if [ "$ch04" -ge 20 ]; then

 continue

else

  echo "The file system is okay now ,please contunie to work!  ^_^ "