Oracle产生redo日志量大小统计

在Oracle中，对于数据库的修改操作都会记录redo，那么不同的操作会产生多少redo呢？可以通过以下一些方式来查询来统计产生的redo日志量。

（1）SQL*Plus中使用AUTOTRACE的使用。

当在SQL*Plus中启用autotrace跟踪后，在执行了特定的DML语句时，Oracle会显示该语句的统计信息，其中，redo Size一栏表示的就是该操作产生的redo的数量，其单位为Bytes：

SCOTT@seiang11g>set autotrace traceonly statistics

注意：如果在启动autotrace跟踪的时候，出现如下报错：SP2-0618: Cannot find the Session Identifier.  Check PLUSTRACE role is enabled.
解决方法请参考另一篇博文：SP2-0618: Cannot find the Session Identifier.Check PLUSTRACE role is enabled

 

SCOTT@seiang11g>create table emp1 as select * from emp;

 

Table created.

 

SCOTT@seiang11g>

SCOTT@seiang11g>insert into emp1 select * from emp1;

 

14 rows created.

 

 

Statistics

----------------------------------------------------------

         15  recursive calls

         22  db block gets

         33  consistent gets

          5  physical reads

       1872  redo size

        834  bytes sent via SQL*Net to client

        791  bytes received via SQL*Net from client

          3  SQL*Net roundtrips to/from client

          2  sorts (memory)

          0  sorts (disk)

         14  rows processed

 

（2）通过v$mystat查询。

Oracle通过v$mystat视图记录当前session的统计信息，我们也可以从该视图中查询得到session的redo生成情况：

SCOTT@seiang11g>set autot off

SCOTT@seiang11g>

SCOTT@seiang11g>select a.name,b.value from v$statname a,v$mystat b

  2  where a.statistic# = b.statistic# and a.name='redo size';

 

NAME                                                                  VALUE

---------------------------------------------------------------- ----------

redo size                                                             29140

 

SCOTT@seiang11g>

SCOTT@seiang11g>insert into emp1 select * from emp1;

 

28 rows created.

 

SCOTT@seiang11g>

SCOTT@seiang11g>select a.name,b.value from v$statname a,v$mystat b

  2  where a.statistic# = b.statistic# and a.name='redo size';

 

NAME                                                                  VALUE

---------------------------------------------------------------- ----------

redo size                                                             30708

 

SCOTT@seiang11g>

SCOTT@seiang11g>select 30708-29140 from dual;

 

30708-29140

-----------

       1568

 

（3）通过v$sysstat查询。
对于数据库全局Redo的生成量，可以通过v$sysstat视图来查询得到：

SYS@seiang11g>select name,value from v$sysstat where name='redo size';

 

NAME                                                                  VALUE

---------------------------------------------------------------- ----------

redo size                                                         548518160

从v$sysstat视图中得到的是自数据库实例启动以来的累积日志生成量，可以根据实例启动时间大致估算每天数据库的日志生成量：

 

SYS@seiang11g>alter session set nls_date_format='yyyy-mm-dd hh34:mi:ss';

 

Session altered.

 

SYS@seiang11g>

SYS@seiang11g>select

  2      (select value/1024/1024/1024 from v$sysstat where name='redo size'

  3       )/

  4      (select round(sysdate-

  5          (select startup_time from v$instance

  6          )) from dual

  7      ) redo_gb_per_day

  8      from dual;

 

REDO_GB_PER_DAY

---------------

     .102173401

 

如果数据库运行在归档模式下，由于其他因素的影响，以上Redo生成量并不代表归档日志的大小，但是可以通过一定的加权提供参考。

至于归档日志的生成量，可以通过v$archived_log视图，根据一段时间的归档日志量进行估算得到。该视图中记录了归档日志的主要信息：

SYS@seiang11g>select name,completion_time,blocks*block_size/1024/1024 MB

  2  from v$archived_log where status = 'A';

 

NAME                                               COMPLETION_TIME             MB

-------------------------------------------------- ------------------- ----------

/u01/app/oracle/arch/arch_1_949237404_8.log        2017-07-13 13:37:10 1.74072266

/u01/app/oracle/arch/arch_1_949237404_9.log        2017-09-13 17:09:40 35.9506836

/u01/app/oracle/arch/arch_1_949237404_10.log       2017-09-13 22:00:47 42.2592773

/u01/app/oracle/arch/arch_1_949237404_11.log       2017-09-14 05:00:33 36.9936523

/u01/app/oracle/arch/arch_1_949237404_12.log       2017-09-14 19:00:36 36.9335938

/u01/app/oracle/arch/arch_1_949237404_13.log       2017-09-15 01:06:21 35.8876953

/u01/app/oracle/arch/arch_1_949237404_14.log       2017-09-15 15:00:10 35.8935547

/u01/app/oracle/arch/arch_1_949237404_15.log       2017-09-15 22:00:37 37.5634766

/u01/app/oracle/arch/arch_1_949237404_16.log       2017-09-16 06:00:28 42.2397461

/u01/app/oracle/arch/arch_1_949237404_17.log       2017-09-16 14:00:16 43.9946289

/u01/app/oracle/arch/arch_1_949237404_18.log       2017-09-16 22:00:25 44.0483398

/u01/app/oracle/arch/arch_1_949237404_19.log       2017-09-17 06:00:25 40.4213867

/u01/app/oracle/arch/arch_1_949237404_20.log       2017-09-17 14:00:25 42.0063477

/u01/app/oracle/arch/arch_1_949237404_21.log       2017-09-17 22:00:28 42.7241211

/u01/app/oracle/arch/arch_1_949237404_22.log       2017-09-18 11:00:07 36.0229492

 

某日全天的日志生成可以通过如下查询计算：

SYS@seiang11g>select trunc(completion_time),

  2        sum(Mb)/1024 DAY_GB

  3      from

  4        (select name,

  5          completion_time,

  6          blocks*block_size/1024/1024 Mb

  7        from v$archived_log

  8        where completion_time between trunc(sysdate)-2 and trunc(sysdate)-1

  9        )

 10    group by trunc(completion_time);

 

TRUNC(COMPLETION_TI     DAY_GB

------------------- ----------

2017-09-16 00:00:00 .127229214

 

最近日期的日志生成统计：

SYS@seiang11g>select trunc(completion_time),

  2        sum(mb)/1024 day_gb

  3      from

  4        (select name,

  5          completion_time,

  6          blocks*block_size/1024/1024 mb

  7        from v$archived_log

  8        )

  9      group by trunc(completion_time);

 

TRUNC(COMPLETION_TI     DAY_GB

------------------- ----------

2017-09-15 00:00:00  .10678196

2017-09-18 00:00:00 .035178661

2017-09-13 00:00:00 .076376915

2017-09-17 00:00:00 .122218609

2017-07-13 00:00:00 .065961361

2017-09-16 00:00:00 .127229214

2017-09-14 00:00:00 .072194576

根据每日归档的生成量，我们也可以反过来估计每日的数据库活动性及周期性，并决定空间分配等问题。

 

拓展：

（一）以下脚本可以用于列出最近Oracle数据库每小时估算的redo重做日志产生量，因为估算数据来源于archivelog的产生量和大小，所以数据是近似值，可供参考:

 

WITH times AS

 (SELECT

   hour_end_time

    FROM (SELECT (TRUNC(SYSDATE, 'HH') + (2 / 24)) - (ROWNUM / 24) hour_end_time

            FROM DUAL

          CONNECT BY ROWNUM <= (1 * 24) + 3),

         v$database

   WHERE log_mode = 'ARCHIVELOG')

SELECT hour_end_time, NVL(ROUND(SUM(size_mb), 3), 0) size_mb, i.instance_name

  FROM(

SELECT hour_end_time, CASE WHEN(hour_end_time - (1 / 24)) > lag_next_time THEN(next_time + (1 / 24) - hour_end_time) * (size_mb / (next_time - lag_next_time)) ELSE 0 END + CASE WHEN hour_end_time < lead_next_time THEN(hour_end_time - next_time) * (lead_size_mb / (lead_next_time - next_time)) ELSE 0 END + CASE WHEN lag_next_time > (hour_end_time - (1 / 24)) THEN size_mb ELSE 0 END + CASE WHEN next_time IS NULL THEN(1 / 24) * LAST_VALUE(CASE WHEN next_time IS NOT NULL AND lag_next_time IS NULL THEN 0 ELSE(size_mb / (next_time - lag_next_time)) END IGNORE NULLS) OVER(

 ORDER BY hour_end_time DESC, next_time DESC) ELSE 0 END size_mb

  FROM(

SELECT t.hour_end_time, arc.next_time, arc.lag_next_time, LEAD(arc.next_time) OVER(

 ORDER BY arc.next_time ASC) lead_next_time, arc.size_mb, LEAD(arc.size_mb) OVER(

 ORDER BY arc.next_time ASC) lead_size_mb

  FROM times t,(

SELECT next_time, size_mb, LAG(next_time) OVER(

 ORDER BY next_time) lag_next_time

  FROM(

SELECT next_time, SUM(size_mb) size_mb

  FROM(

SELECT DISTINCT a.sequence#, a.next_time, ROUND(a.blocks * a.block_size / 1024 / 1024) size_mb

  FROM v$archived_log a,(

SELECT

CASE WHEN TO_NUMBER(pt.VALUE) = 0 THEN 1 ELSE TO_NUMBER(pt.VALUE) END VALUE

  FROM v$parameter pt

 WHERE pt.name = 'thread') pt

 WHERE a.next_time > SYSDATE - 3 AND a.thread# = pt.VALUE AND ROUND(a.blocks * a.block_size / 1024 / 1024) > 0)

 GROUP BY next_time)) arc

 WHERE t.hour_end_time = (TRUNC(arc.next_time(+), 'HH') + (1 / 24)))

 WHERE hour_end_time > TRUNC(SYSDATE, 'HH') - 1 - (1 / 24)), v$instance i

 WHERE hour_end_time <= TRUNC(SYSDATE, 'HH')

 GROUP BY hour_end_time, i.instance_name

 ORDER BY hour_end_time

 /

 

执行结果：

HOUR_END_TIME          SIZE_MB INSTANCE_NAME

------------------- ---------- ----------------

2017-09-17 14:00:00       5.25 seiang11g

2017-09-17 15:00:00      5.374 seiang11g

2017-09-17 16:00:00      5.374 seiang11g

2017-09-17 17:00:00      5.374 seiang11g

2017-09-17 18:00:00      5.374 seiang11g

2017-09-17 19:00:00      5.374 seiang11g

2017-09-17 20:00:00      5.374 seiang11g

2017-09-17 21:00:00      5.374 seiang11g

2017-09-17 22:00:00      5.374 seiang11g

2017-09-17 23:00:00       2.79 seiang11g

2017-09-18 00:00:00       2.77 seiang11g

2017-09-18 01:00:00       2.77 seiang11g

2017-09-18 02:00:00       2.77 seiang11g

2017-09-18 03:00:00       2.77 seiang11g

2017-09-18 04:00:00       2.77 seiang11g

2017-09-18 05:00:00       2.77 seiang11g

2017-09-18 06:00:00       2.77 seiang11g

2017-09-18 07:00:00       2.77 seiang11g

2017-09-18 08:00:00       2.77 seiang11g

2017-09-18 09:00:00       2.77 seiang11g

2017-09-18 10:00:00       2.77 seiang11g

2017-09-18 11:00:00       2.77 seiang11g

2017-09-18 12:00:00       .005 seiang11g

2017-09-18 13:00:00          0 seiang11g

2017-09-18 14:00:00          0 seiang11g

 

 

（二）Oracle查询最近几天每小时归档日志产生数量的脚本，脚本内容如下所示：

 

SELECT SUBSTR(TO_CHAR(first_time, 'MM/DD/RR HH:MI:SS'),1,5) Day,

    SUM(DECODE(SUBSTR(TO_CHAR(first_time, 'MM/DD/RR HH24:MI:SS'),10,2),'00',1,0)) H00,

    SUM(DECODE(SUBSTR(TO_CHAR(first_time, 'MM/DD/RR HH24:MI:SS'),10,2),'01',1,0)) H01,

    SUM(DECODE(SUBSTR(TO_CHAR(first_time, 'MM/DD/RR HH24:MI:SS'),10,2),'02',1,0)) H02,

    SUM(DECODE(SUBSTR(TO_CHAR(first_time, 'MM/DD/RR HH24:MI:SS'),10,2),'03',1,0)) H03,

    SUM(DECODE(SUBSTR(TO_CHAR(first_time, 'MM/DD/RR HH24:MI:SS'),10,2),'04',1,0)) H04,

    SUM(DECODE(SUBSTR(TO_CHAR(first_time, 'MM/DD/RR HH24:MI:SS'),10,2),'05',1,0)) H05,

    SUM(DECODE(SUBSTR(TO_CHAR(first_time, 'MM/DD/RR HH24:MI:SS'),10,2),'06',1,0)) H06,

    SUM(DECODE(SUBSTR(TO_CHAR(first_time, 'MM/DD/RR HH24:MI:SS'),10,2),'07',1,0)) H07,

    SUM(DECODE(SUBSTR(TO_CHAR(first_time, 'MM/DD/RR HH24:MI:SS'),10,2),'08',1,0)) H08,

    SUM(DECODE(SUBSTR(TO_CHAR(first_time, 'MM/DD/RR HH24:MI:SS'),10,2),'09',1,0)) H09,

    SUM(DECODE(SUBSTR(TO_CHAR(first_time, 'MM/DD/RR HH24:MI:SS'),10,2),'10',1,0)) H10,

    SUM(DECODE(SUBSTR(TO_CHAR(first_time, 'MM/DD/RR HH24:MI:SS'),10,2),'11',1,0)) H11,

    SUM(DECODE(SUBSTR(TO_CHAR(first_time, 'MM/DD/RR HH24:MI:SS'),10,2),'12',1,0)) H12,

    SUM(DECODE(SUBSTR(TO_CHAR(first_time, 'MM/DD/RR HH24:MI:SS'),10,2),'13',1,0)) H13,

    SUM(DECODE(SUBSTR(TO_CHAR(first_time, 'MM/DD/RR HH24:MI:SS'),10,2),'14',1,0)) H14,

    SUM(DECODE(SUBSTR(TO_CHAR(first_time, 'MM/DD/RR HH24:MI:SS'),10,2),'15',1,0)) H15,

    SUM(DECODE(SUBSTR(TO_CHAR(first_time, 'MM/DD/RR HH24:MI:SS'),10,2),'16',1,0)) H16,

    SUM(DECODE(SUBSTR(TO_CHAR(first_time, 'MM/DD/RR HH24:MI:SS'),10,2),'17',1,0)) H17,

    SUM(DECODE(SUBSTR(TO_CHAR(first_time, 'MM/DD/RR HH24:MI:SS'),10,2),'18',1,0)) H18,

    SUM(DECODE(SUBSTR(TO_CHAR(first_time, 'MM/DD/RR HH24:MI:SS'),10,2),'19',1,0)) H19,

    SUM(DECODE(SUBSTR(TO_CHAR(first_time, 'MM/DD/RR HH24:MI:SS'),10,2),'20',1,0)) H20,

    SUM(DECODE(SUBSTR(TO_CHAR(first_time, 'MM/DD/RR HH24:MI:SS'),10,2),'21',1,0)) H21,

    SUM(DECODE(SUBSTR(TO_CHAR(first_time, 'MM/DD/RR HH24:MI:SS'),10,2),'22',1,0)) H22,

    SUM(DECODE(SUBSTR(TO_CHAR(first_time, 'MM/DD/RR HH24:MI:SS'),10,2),'23',1,0)) H23,

    COUNT(*) TOTAL

FROM v$log_history a

WHERE first_time>=to_char(sysdate-10)

GROUP BY SUBSTR(TO_CHAR(first_time, 'MM/DD/RR HH:MI:SS'),1,5)

ORDER BY SUBSTR(TO_CHAR(first_time, 'MM/DD/RR HH:MI:SS'),1,5) DESC;

 

修改天数，可以修改WHERE first_time>=to_char(sysdate-11) 

 

执行结果：

参考链接：

http://www.dbtan.com/2009/12/how-many-redo-has-produced.html

http://www.askmaclean.com/archives/script%E5%88%97%E5%87%BAoracle%E6%AF%8F%E5%B0%8F%E6%97%B6%E7%9A%84redo%E9%87%8D%E5%81%9A%E6%97%A5%E5%BF%97%E4%BA%A7%E7%94%9F%E9%87%8F.html

http://www.jb51.net/article/119200.htm

作者：SEian.G（苦练七十二变，笑对八十一难）