IIS日志分析

IIS日志分析工具：LogParse、sqlserver 2008

1.从将IIS日志转换到数据库表中

logparser "SELECT TO_LOCALTIME(TO_TIMESTAMP(ADD(TO_STRING(date, 'yyyy-MM-dd '), TO_STRING(time, 'hh:mm:ss')),'yyyy-MM-dd hh:mm:ss')) AS RequestTime, * FROM 'D:phone.log' to phonelog" -i:IISW3C -o:SQL -oConnString:"Driver={SQL Server};server=.;database=iis;uid=sa;pwd=pwd;Integrated Security=SSPI" -createtable:ON

2.使用sql分析语句分析结果

--去除url参数并保存到指定表中
alter table phonelog add url nvarchar(200)
go
update phonelog set url=case when Charindex('/',csUriStem,CHARINDEX('/',csUriStem,2)+1)>0 then SUBSTRING(csUriStem,0,Charindex('/',csUriStem,CHARINDEX('/',csUriStem,2)+1)) else csUriStem end

--分析时间占用
select scStatus, count(*) AS count, sum(timetaken * 1.0) /1000.0 AS sum_timetaken_second from phonelog with(nolock) group by scStatus order by 3 desc

--统计网络级别错误
declare @recCount bigint;
select @recCount = count(*) from phonelog with(nolock)
select case when scWin32Status=64 then '客户端连接已关闭' when scWin32Status=121 then '传输超时' when scWin32Status=1236 then '本地网络中断' else Convert(varchar,scWin32Status) end as scWin32Status , count(*) AS count, (count(*) * 100.0 / @recCount) AS [percent]
from phonelog with(nolock)
where scWin32Status > 0
group by scWin32Status
order by 2 desc

--查询每个用户的最大最小请求次数
select cIp,COUNT(*) as count from phonelog group by cIp order by 2 desc

--查询日志最大最小时间
select MAX(l.date) as maxdate,MIN(l.date) as mindate from phonelog l

--查询耗时最大的请求
select top 1000 l.csUriStem,(timetaken * 1.0) /1000.0 as timetaken_second from phonelog l order by l.timeTaken desc

--查询标准请求中请求最多的url
select top 1000 l.csUriStem,sum(timetaken * 1.0) /1000.0 as sum_timetaken_second,COUNT(*) as count from phonelog l group by l.csUriStem order by 3 desc

--查询发杂请求中请求最多的接口
select ll.url,ll.count,ll.sum_timetaken_second,ll.sum_timetaken_second/ll.count as avg_timetaken_second from (
select l.url,sum(timetaken * 1.0) /1000.0 as sum_timetaken_second,COUNT(*) as count from phonelog l group by l.url)ll order by 2 desc

--查询平均耗时最大的请求接口
select ll.url,ll.count,ll.sum_timetaken_second,ll.sum_timetaken_second/ll.count as avg_timetaken_second from (
select l.url,sum(timetaken * 1.0) /1000.0 as sum_timetaken_second,COUNT(*) as count from phonelog l group by l.url)ll order by 4 desc


select * from phonelog l where l.csuristem like '%GetClassDynamicPager%'

3.其它数据库操作和分析可以随自己的意愿去写sql语句。

4.根据分析结果和业务，查找性能瓶颈，有针对的对当前的程序进行优化修改。