Chinaunix首页 | 论坛 | 博客
  • 博客访问: 511236
  • 博文数量: 161
  • 博客积分: 6010
  • 博客等级: 准将
  • 技术积分: 1947
  • 用 户 组: 普通用户
  • 注册时间: 2007-08-25 01:20
文章分类

全部博文(161)

文章存档

2011年(44)

2010年(47)

2009年(48)

2008年(22)

我的朋友

分类: Oracle

2011-03-02 12:52:49

与undo 有关的事件
SELECT * FROM v$statname WHERE NAME LIKE '%undo%' 
---------- -------------------------------------------
DBWR undo block writes                                                                 
undo change vector size                                                                
transaction tables consistent reads - undo records applied                             
data blocks consistent reads - undo records applied                                    
rollback changes - undo records applied                                                
auto extends on undo tablespace                                                        
total number of undo segments dropped                                                  
global undo segment hints helped                                                       
global undo segment hints were stale                                                   
local undo segment hints helped                                                        
local undo segment hints were stale                                                    
undo segment header was pinned                                                         
SMON posted for undo segment recovery                                                  
SMON posted for undo segment shrink                                                    
IMU undo retention flush                                                               
IMU undo allocation size                                                               

---------------------------------------------------------------------
比较关注的统计项目
undo change vector size
产生undo数据的大小,bytes计算,分析单个事务|整个系统的
undo产生变化情况都是很有用

data blocks consistent reads --undo records applied
一致性读,bytes

rollback changes --undo  reocords applied
回滚操作所应用的undo记录的大小,单位是bytes
----------------------------------------------------------------------

--------------------实验- --------------------------
--mystat.sql
set verify off
column value new_val V
define S="&1"
set autotrace off
select a.name, b.value
from v$statname a, v$mystat b
where a.statistic# = b.statistic#
and lower(a.name) like '%' || lower('&S')||'%'
/


-- mystat2.sql
set verify off
select a.name, b.value V, to_char(b.value-&V,'999,999,999,999') diff
from v$statname a, v$mystat b
where a.statistic# = b.statistic#
and lower(a.name) like '%' || lower('&S')||'%'
/
---通过观察v$mystat变换来观察当前会话事务undo的产生情况



--表DBA_hist_sysstat 每隔1小时的v$sysstat的快照,
--通用,将undi change 改成别的事件名称
SET pages 999
SET lines 140
col curval FORMAT 99999999999999999
col prevval FORMAT 99999999999999999
SELECT cur_stat.snap_id CurID,
       prev_stat.snap_id PrevID,
       To_char(sn.begin_interval_time, 'YYYY-MM-DD HH24:MI') BeginTime,
       To_char(sn.end_interval_time, 'YYYY-MM-DD HH24:MI') EndTime,
       cur_stat.VALUE                       curval,
       prev_stat.VALUE                      prevval,
       ROUND(( cur_stat.VALUE - prev_stat.VALUE )/1024/1024,2) "UNDO Gen(MB/Hour)"
FROM   dba_hist_snapshot sn,
       (SELECT snap_id, VALUE
        FROM   dba_hist_sysstat
        WHERE  stat_name = 'undo change vector size') cur_stat,
       (SELECT snap_id,VALUE
        FROM   dba_hist_sysstat
        WHERE  stat_name = 'undo change vector size') prev_stat
WHERE  sn.snap_id = cur_stat.snap_id
       AND cur_stat.snap_id = prev_stat.snap_id + 1
       AND sn.begin_interval_time > sysdate-3
ORDER  BY 1
/

--after run
CURID PREVID BEGINTIME        ENDTIME             CURVAL    PREVVAL UNDO Gen(MB/Hour)
----- ------ ---------------- ---------------- --------- ---------- -----------------
    3      2 2011-02-27 02:56 2011-02-27 03:07    188080    1298732             -1.06
    4      3 2011-02-27 03:07 2011-02-27 04:00    656708     188080               .45
    5      4 2011-02-27 04:00 2011-02-27 05:00   1122792     656708               .44
    6      5 2011-02-27 05:00 2011-02-27 06:00   1547508    1122792               .41
    7      6 2011-02-27 06:00 2011-02-27 07:00   1976980    1547508               .41
    8      7 2011-02-28 23:57 2011-03-01 00:08   6570100    1976980              4.38
    9      8 2011-03-01 00:08 2011-03-01 01:00  12689980    6570100              5.84
   10      9 2011-03-01 01:00 2011-03-01 02:00  13180860   12689980               .47
   11     10 2011-03-01 02:00 2011-03-01 03:00  15099052   13180860              1.83
   12     11 2011-03-01 12:08 2011-03-01 12:19    140304   15099052            -14.27
   13     12 2011-03-01 12:19 2011-03-01 13:00    589088     140304               .43
   14     13 2011-03-01 13:00 2011-03-01 14:00   1241508     589088               .62
   15     14 2011-03-01 22:09 2011-03-01 22:20   3406144    1241508              2.06
   
------------------------------------------------------------------------------------
--显示的UNDO表空间说明当前正在被使用,不显示说明没被使用
--(一)
col tablespace_name for a20
SELECT D.TABLESPACE_NAME,
D.SPACE "TOTAL_SPACE(M)",
U.U_SPACE "USED_SPACE(M)",
ROUND((U.U_SPACE/D.SPACE)*100,2) "USED_RATE(%)"
FROM 
(SELECT TABLESPACE_NAME,ROUND(SUM(BYTES)/(1024*1024),2) SPACE
FROM DBA_DATA_FILES
GROUP BY TABLESPACE_NAME) D,
(SELECT TABLESPACE_NAME,SUM(BYTES)/1024/1024 U_SPACE  FROM DBA_UNDO_EXTENTS WHERE STATUS='ACTIVE' 
GROUP BY TABLESPACE_NAME) U
WHERE D.TABLESPACE_NAME=U.TABLESPACE_NAME;
--after run


TABLESPACE_NAME      TOTAL_SPACE(M) USED_SPACE(M) USED_RATE(%)
-------------------- -------------- ------------- ------------
UNDOTBS1                         25         .0625          .25

--详细(二)
SELECT seg.tablespace_name "Tablespace Name", ts.bytes/1024/1024 "TS Size(MB)",
  ue.status "UNDO Status", count(*) "Used Extents",
  round(sum(ue.bytes)/1024/1024, 2) "Used Size(MB)",
  round(sum(ue.bytes)/ts.bytes*100, 2) "Used Rate(%)"
FROM dba_segments seg, DBA_UNDO_EXTENTS ue,
  (SELECT tablespace_name, sum(bytes) bytes
    FROM dba_data_files GROUP BY tablespace_name) ts
WHERE ue.segment_NAME=seg.segment_NAME and seg.tablespace_name=ts.tablespace_name
GROUP BY seg.tablespace_name, ts.bytes, ue.status
ORDER BY seg.tablespace_name;
--after run
Tablespace TS Size(MB) UNDO Stat Used Extents Used Size(MB) Used Rate(%)
---------- ----------- --------- ------------ ------------- ------------
UNDOTBS1            25 ACTIVE               1           .06          .25
UNDOTBS1            25 EXPIRED            122         13.25           53
UNDOTBS1            25 UNEXPIRED           13           .81         3.25
 
--查找session,undo 大小
--等等
--得到活动事务的UNDO,sid,得到的语句并不准确
--是当前的SID所执行的最后一个SQL语句
col sid for 99999
col osuser for a8
col username for a12
col tablespace_name for a10
col segment for a15
col sql_text for a30
col status for a10
set lines 150 pages 999
SELECT sess.SID,
sess.serial#, sess.osuser, sess.username,
rseg.segment_name SEGMENT, rseg.tablespace_name,
trans.status,
trans.used_ublk,
trans.used_ublk*8/1024 "UNDO SIZE(M)",
rseg.STATUS, 
sa.sql_text,
sess.program
FROM v$session sess, v$transaction trans, dba_rollback_segs rseg, v$sql sa
WHERE sess.taddr=trans.addr AND trans.xidusn=rseg.segment_id(+)
AND (sess.sql_hash_value=sa.hash_value OR sess.prev_hash_value=sa.hash_value)
ORDER BY sql_text;


阅读(973) | 评论(2) | 转发(0) |
给主人留下些什么吧!~~

chinaunix网友2011-06-05 01:47:50

大连法律咨询在线 http://www.fabowang.com 大连律师在线咨询 http://www.fabowang.com 大连法律顾问网 http://www.fabowang.com 大连律师咨询 http://www.fabowang.com

chinaunix网友2011-03-06 16:36:38

很好的, 收藏了 推荐一个博客,提供很多免费软件编程电子书下载: http://free-ebooks.appspot.com