Crawl specific fields in the Log shell script

Boss let me put the information in the log summary, good trouble, wrote this script to help work.

Need to improve: 1. Crawl log time, only the last time, and this script will catch

2. Requires two time difference (e.g. 15:23:34 and 23:23:34), not completed

3.aix don't get used to it

#!/bin/ksh# Enter the destination file directory cd /user/lyb# Enter the identification field echo  start time     end time     customer group      number of customers  >> /user/file10#for i in  1 2 3 4  5 6 7 8 9 10 11 12 "while [ true ]doecho " please  input the filename "#变量名为文件名, Read -p does not make read filename  #取出需要的字段, when there are multiple identical fields in the file, Without judging the conditions s1= ' grep   ' received the registration Response message packet, start performing task resolution "  $filename   | sed  ' s/\. *//g ' s2= ' grep   "terminate the task processing finished, service exits"     $filename  | sed  ' s/\. *//g "s3= ' grep  " loads the customer group. * Users   $filename  | sed  ' s/^.*\[//g '  | sed  ' s /\].*$//g ' s4= ' grep   "loaded. * users"    $filename  | sed  ' s/^.*\[//g '  |  sed  ' s/\].*$//g ' #将取出的字段输入到文件echo      "$s 1"         " $s 2 "       " $s 3"         " $s 4 " >> /user/file10 done 

#计算时间差sql语句

#select Round ((to_date (' 2014-11-13 12:22:23 ', ' yyyy-mm-dd hh24:mi:ss ')-to_date (' 2014-11-13 12:12:32 ', ' yyyy-mm-dd Hh24:mi:ss ') *24*60) time from dual;

Crawl specific fields in the Log shell script