查找bad sql的方法

查找运行系统里bad sql是一个古老的话题, 我们要根据自己的实际情况来分析。

绝不能教条的运用下面介绍的这些方法。

使用这些SQL语句时，会对系统表产生分组操作，当然也增大了系统的负载。

建议大家在系统启动了一段时间后，在半夜负载较轻的时间定时(例如:一个月)来查一查。一定要具体问题具体分析。

下面是我收藏的一些查找bad sql的方法:

column sql_text format a80;
-- 值得怀疑的SQL 来自http://www.ixora.com.au/
select
substr(to_char(s.pct, @#99.00@#), 2) || @#%@# load,
s.executions executes,
p.sql_text
from
(
    select
      address,
      disk_reads,
      executions,
      pct,
      rank() over (order by disk_reads desc) ranking
    from
      (
        select
          address,
          disk_reads,
          executions,
          100 * ratio_to_report(disk_reads) over () pct
        from
          sys.v_$sql
        where
          command_type != 47
      )
    where
      disk_reads > 50 * executions
) s,
sys.v_$sqltext p
where
s.ranking <= 5 and
p.address = s.address
order by
1, s.address, p.piece
/

-- 逻辑读多的SQL
select * from (select buffer_gets, sql_text
from v$sqlarea
where buffer_gets > 500000
order by buffer_gets desc) where rownum<=30;

-- 执行次数多的SQL
select sql_text,executions from
(select sql_text,executions from v$sqlarea order by executions desc)
where rownum<81;

-- 读硬盘多的SQL
select sql_text,disk_reads from
(select sql_text,disk_reads from v$sqlarea order by disk_reads desc)
where rownum<21;

-- 排序多的SQL
select sql_text,sorts from
(select sql_text,sorts from v$sqlarea order by sorts desc)
where rownum<21;

--分析的次数太多，执行的次数太少，要用绑变量的方法来写sql
set pagesize 600;
set linesize 120;
select substr(sql_text,1,80) "sql", count(*), sum(executions) "totexecs"
   from v$sqlarea
   where executions < 5
   group by substr(sql_text,1,80)
   having count(*) > 30
   order by 2;

-- 游标的观察
set pages 300;
select sum(a.value), b.name
     from v$sesstat a, v$statname b
     where a.statistic# = b.statistic#
       and b.name = @#opened cursors current@#
     group by b.name;

select count(0) from v$open_cursor;

select user_name,sql_text,count(0) from v$open_cursor
group by user_name,sql_text having count(0)>30;

--查看当前用户&username执行的SQL
select sql_text from v$sqltext_with_newlines where (hash_value,address) in
(select sql_hash_value,sql_address from v$session where username=@#&username@#)
order by address,piece;

下一篇：给SQL Server传送数组参数的变通办法

↓相关文章：