vc-mysql-sniffer统计MySQL的SQL分布

有时候我们需要统计线上的SQL执行情况，比如想知道哪条SQL执行最频繁，我们可以开启general_log，然后进行统计，但是general_log开启非常损耗性能，那么我们可以使用vc-mysql-sniffer来代替，该工具是编译好的二进制，下载即可使用。下载地址：

https://www.vividcortex.com/resources/network-analyzer-for-mysql

我这里修改了网上的一个脚本，分析vc-mysql-sniffer抓取到的结果统计频繁执行的SQL。

#!/usr/bin/python
#coding:utf8
# python analysis-vc-log.py 3315 | sort | uniq -c | sort -nr |head -n 10
import re
import sys
import os
import commands
vc_sniffer_time=5
port=sys.argv[1]
vc_cmd=""" /usr/bin/timeout %s  /data/software/vc-mysql-sniffer -binding="[::]:%s"  > /tmp/tmp_vc_mysql_%s.txt """ % (vc_sniffer_time,port,port)
outtext = commands.getoutput(vc_cmd)
cmd=""" grep -Ev '# Time:|# User@Host' /tmp/tmp_vc_mysql_%s.txt |sed 's/# Query_time.*/myxxxxx/g' |awk BEGIN{RS=EOF}'{gsub(/\n/," ");print}'|awk BEGIN{RS=EOF}'{gsub(/myxxxxx/,"\n");print}' >/tmp/vc_mysql_%s.txt""" % (port,port)
outtext = commands.getoutput(cmd)
file="/tmp/vc_mysql_%s.txt" % (port)
logFo = open(file)
for line in logFo:
    line = re.sub(r"
","",line)
    lineMatch = re.match(r".*",line)
    if lineMatch:
        lineTmp = lineMatch.group(0)
        lineTmp = lineTmp.lower()
        # remove extra space
        lineTmp = re.sub(r"s+", " ",lineTmp)
        # replace values (value) to values (x)
        lineTmp = re.sub(r"valuess*(.*?)", "values (x)",lineTmp)
        # replace filed = 'value' to filed = 'x'
        lineTmp = re.sub(r"(=|>|<|>=|<=)s*('|").*?2","\1 'x'",lineTmp)
        # replace filed = value to filed = x
        lineTmp = re.sub(r"(=|>|<|>=|<=)s*[0-9]+","\1 x",lineTmp)
        # replace like 'value' to like 'x'
        lineTmp = re.sub(r"likes+('|").*?1","like 'x'",lineTmp)
        # replace in (value) to in (x)
        lineTmp = re.sub(r"ins+(.*?)","in (x)",lineTmp)
        # replace between '...' and '...' to between 'x' and 'x'
        lineTmp = re.sub(r"betweens+('|").*?1s+ands+1.*?1","between 'x' and 'x' ",lineTmp)
        # replace between ... and ... to between x and x
        lineTmp = re.sub(r"betweens+[0-9]+s+ands+[0-9]+","between x and x ",lineTmp)
        # replace limit x,y to limit
        lineTmp = re.sub(r"limit.*","limit",lineTmp)
        print lineTmp
logFo.close()

python analysis-vc-log.py 3310 | sort | uniq -c | sort -nr |head -n 10

抓取mysql 3310端口，统计执行最频繁的10条SQL。结果我这里就不贴上来了。有兴趣的同学自己测试。

参考资料：

https://www.centos.bz/2015/02/analysis-mysql-general-log/