这篇文章主要介绍了查找Oracle高消耗语句的方法,需要的朋友可以参考下

在运行下面的脚本之前需要先用生成AWR报告的SQL（程序脚本一般保存在$ORACLE_HOME下的rdbms/admin中，名称为awrrpt.sql，需要输入生成AWR报告的天数范围）找到开始和结束的snapshot编号：begin_snap和end_snap。  代码如下: <span style="font-size:18px;">set line 1000  set linesize 200  set pagesize 2000  set long 999999  set echo on  set markup html on  select res.*  from (select to_char(d.end_interval_time,'yyyy-mm-dd'),  a.PARSING_SCHEMA_NAME,  c.MODULE,  a.sql_id,  a.execs as 执行次数,  ROUND(a.cpu_times / a.execs, 2) as 单次执行时间,  a.cpu_times as cpu消耗时间,  ROUND(a.cpu_times / b.sum_time * 100, 2) as 消耗cpu百分比,  a.buffer_gets as 逻辑读,  ROUND(a.buffer_gets / b.sum_buffer * 100, 2) as 逻辑读百分比,  a.disk_read as 物理读,  ROUND(a.disk_read / b.sum_disk * 100, 2) as 物理读百分比,  c.sql_fulltext  from (select PARSING_SCHEMA_NAME,  sql_id,  sum(EXECUTIONS_DELTA) AS execs,  round(sum(CPU_TIME_DELTA) / 1000000, 2) AS cpu_times,  round(sum(ELAPSED_TIME_DELTA) / 1000000, 2) AS elapsed_time,  sum(BUFFER_GETS_DELTA) AS buffer_gets,  sum(DISK_READS_DELTA) AS disk_read  from sys.WRH$_SQLSTAT wr, gv$instance i  where SNAP_ID <= &end_snap  and snap_id >= &begin_snap  and wr.INSTANCE_NUMBER = i.INSTANCE_NUMBER  and i.instance_number = &instance_number  group by PARSING_SCHEMA_NAME, wr.INSTANCE_NUMBER, sql_id) a,  (SELECT round(SUM(CPU_TIME_DELTA) / 1000000, 2) sum_time,  SUM(BUFFER_GETS_DELTA) sum_buffer,  sum(DISK_READS_DELTA) sum_disk  FROM sys.WRH$_SQLSTAT wr, gv$instance i  where SNAP_ID <= &end_snap  and snap_id >= &begin_snap  and wr.INSTANCE_NUMBER = i.INSTANCE_NUMBER  and i.instance_number = &instance_number) b,  v$sqlarea c,  dba_hist_snapshot d  where a.execs > 0  and a.sql_id = c.sql_id  and a.PARSING_SCHEMA_NAME <> 'SYS'  and d.snap_id = &end_snap  order by cpu消耗时间 desc) res  where rownum < 41;  exit</span>    将脚本输出内容保存到记事本txt中，并将记事本的后缀名更改为.html，这样就可以输出以下的网页内容：