shell的思路决定效率

目的：
  运用shell写一个脚本来分析apache日志。按照访问apache的IP地址最多的IP进行排序，并统计最多的10个IP的访问次数。

第一个方法：
[root@trffic root]# du -h /data/logs/db440/colorme-access_log.20060824
637M  /data/logs/db440/colorme-access_log.20060824
[root@trffic root]# time ./filter.sh /data/logs/db440/colorme-access_log.20060824
1    11000  222.191.76.41
2    7882  222.84.8.206
3    6144  60.25.70.41
4    4546  61.54.148.199
5    4194  218.65.239.74
6    3968  218.247.53.164
7    3026  59.37.115.121
8    2878  218.22.141.6
9    2792  219.154.230.18
10    2676  220.198.68.132

real  0m35.587s
user  0m18.130s
sys   0m1.620s
[root@trffic root]#cat filter.sh(内容请详见以下链接)

#!/bin/sh
# Author: jackylau 

USAGE="Usage: $0 /path/to/filename"

if [ $# -ne  1 ];then
        echo $USAGE
        exit
fi

if [ ! -d /tmp/temp ];then
        mkdir /tmp/temp
fi

awk '{print $1}' $1>/tmp/temp/address

cat /tmp/temp/address|sort|uniq -c|sort -nr|head|awk '{print NR"\t",$1"\t",$2}'


rm -rf /tmp/temp

第二种方法：
一个小时也没统计出来
[root@trffic root]#cat filter2.sh(内容请详见以下链接)

#!/bin/sh
# Author: jackylau 

USAGE="Usage: $0 /path/to/filename"

if [ $# -ne  1 ];then
        echo $USAGE
        exit
fi

if [ ! -d /tmp/temp ];then
        mkdir /tmp/temp
fi

awk '{print $1}' $1 >/tmp/temp/address

Count=`cat /tmp/temp/address|sort|uniq|wc -l`
for (( i=0; i<="$Count"; i++ ))
do
        for IP in `head -n 1 /tmp/temp/address`
        do      
                echo "$IP        `awk '{print $1}' $1|grep -c $IP`" >>/tmp/temp/after_address
                echo "wq"|ex -c "g/$IP/d" /tmp/temp/address
        done

done
cat /tmp/temp/after_address|sort -n -r +1|head

rm -rf /tmp/temp