alter table move跟shrink space的區別（轉）
alter table move跟shrink space的區別
今天主要從兩點說他們的區別：
1. 碎片的整理
2.空間的收縮
SQL> select * from v$version;
BANNER
----------------------------------------------------------------
Oracle Database 10g Enterprise Edition Release 10.2.0.1.0 - 64bi
PL/SQL Release 10.2.0.1.0 - Production
CORE 10.2.0.1.0 Production
TNS for Linux: Version 10.2.0.1.0 - Production
NLSRTL Version 10.2.0.1.0 - Production
SQL>
建立測試表
SQL> create table test3 as
2 select rownum id,
3 dbms_random.string('a', round(dbms_random.value(0,10))) col1,
4 trunc(sysdate) - dbms_random.value(1, 365*2) col2
5 from dual connect by rownum<=10000;
Table created
SQL> select count(1) from test3;
COUNT(1)
----------
10000
檢視錶test3的blocks使用情況：
SQL> exec show_space_t('TEST3','auto','table','Y');
Total Blocks............................40
Total Bytes.............................327680
Unused Blocks...........................3
Unused Bytes............................24576
Last Used Ext FileId....................31
Last Used Ext BlockId...................481921
Last Used Block.........................5
*************************************************
0% -- 25% free space blocks.............0
0% -- 25% free space bytes..............0
25% -- 50% free space blocks............0
25% -- 50% free space bytes.............0
50% -- 75% free space blocks............0
50% -- 75% free space bytes.............0
75% -- 100% free space blocks...........0
75% -- 100% free space bytes............0
Unused Blocks...........................0
Unused Bytes............................0
Total Blocks............................32
Total bytes.............................262144
PL/SQL procedure successfully completed
製造碎片
SQL> DELETE FROM TEST3 WHERE MOD(ID,3)=1;
3334 rows deleted
SQL> commit;
Commit complete
發現有碎片了
SQL> exec show_space_t('TEST3','auto','table','Y');
Total Blocks............................40
Total Bytes.............................327680
Unused Blocks...........................3
Unused Bytes............................24576
Last Used Ext FileId....................31
Last Used Ext BlockId...................481921
Last Used Block.........................5
*************************************************
0% -- 25% free space blocks.............0
0% -- 25% free space bytes..............0
25% -- 50% free space blocks............31
25% -- 50% free space bytes.............253952
50% -- 75% free space blocks............1
50% -- 75% free space bytes.............8192
75% -- 100% free space blocks...........0
75% -- 100% free space bytes............0
Unused Blocks...........................0
Unused Bytes............................0
Total Blocks............................0
Total bytes.............................0
PL/SQL procedure successfully completed
SQL>
消除碎片
SQL> alter table test3 move;
Table altered
檢視碎片消除的效果
SQL> exec show_space_t('TEST3','auto','table','Y');
Total Blocks............................32
Total Bytes.............................262144
Unused Blocks...........................6
Unused Bytes............................49152
Last Used Ext FileId....................31
Last Used Ext BlockId...................485065
Last Used Block.........................2
*************************************************
0% -- 25% free space blocks.............0
0% -- 25% free space bytes..............0
25% -- 50% free space blocks............0
25% -- 50% free space bytes.............0
50% -- 75% free space blocks............0
50% -- 75% free space bytes.............0
75% -- 100% free space blocks...........0
75% -- 100% free space bytes............0
Unused Blocks...........................0
Unused Bytes............................0
Total Blocks............................22
Total bytes.............................180224
PL/SQL procedure successfully completed
SQL>
從以上看，碎片整理的效果很好！！！
下面在測試用shrink整理碎片
重建測試環境
SQL> drop table test3;
Table dropped
SQL>
SQL> create table test3 as
2 select rownum id,
3 dbms_random.string('a', round(dbms_random.value(0,10))) col1,
4 trunc(sysdate) - dbms_random.value(1, 365*2) col2
5 from dual connect by rownum<=10000;
Table created
檢視test3的blocks的使用
SQL> exec show_space_t('TEST3','auto','table','Y');
Total Blocks............................40
Total Bytes.............................327680
Unused Blocks...........................3
Unused Bytes............................24576
Last Used Ext FileId....................31
Last Used Ext BlockId...................481921
Last Used Block.........................5
*************************************************
0% -- 25% free space blocks.............0
0% -- 25% free space bytes..............0
25% -- 50% free space blocks............0
25% -- 50% free space bytes.............0
50% -- 75% free space blocks............0
50% -- 75% free space bytes.............0
75% -- 100% free space blocks...........0
75% -- 100% free space bytes............0
Unused Blocks...........................0
Unused Bytes............................0
Total Blocks............................32
Total bytes.............................262144
PL/SQL procedure successfully completed
製造碎片
SQL> delete from test3 where mod(id,3)=1;
3334 rows deleted
SQL> commit;
Commit complete
檢視碎片情況
SQL> exec show_space_t('TEST3','auto','table','Y');
Total Blocks............................40
Total Bytes.............................327680
Unused Blocks...........................3
Unused Bytes............................24576
Last Used Ext FileId....................31
Last Used Ext BlockId...................481921
Last Used Block.........................5
*************************************************
0% -- 25% free space blocks.............0
0% -- 25% free space bytes..............0
25% -- 50% free space blocks............31
25% -- 50% free space bytes.............253952
50% -- 75% free space blocks............0
50% -- 75% free space bytes.............0
75% -- 100% free space blocks...........0
75% -- 100% free space bytes............0
Unused Blocks...........................0
Unused Bytes............................0
Total Blocks............................1
Total bytes.............................8192
PL/SQL procedure successfully completed
用oracle10g新功能整理碎片
SQL> alter table test3 shrink space compact cascade;
alter table test3 shrink space compact cascade
ORA-10636: ROW MOVEMENT is not enabled
SQL> alter table test3 enable row movement;
Table altered
SQL> alter table test3 shrink space compact cascade;
Table altered
再次檢視碎片的情況，發現還有一些碎片，整理碎片效果不好
SQL> exec show_space_t('TEST3','auto','table','Y');
Total Blocks............................40
Total Bytes.............................327680
Unused Blocks...........................3
Unused Bytes............................24576
Last Used Ext FileId....................31
Last Used Ext BlockId...................481921
Last Used Block.........................5
*************************************************
0% -- 25% free space blocks.............1
0% -- 25% free space bytes..............8192
25% -- 50% free space blocks............2
25% -- 50% free space bytes.............16384
50% -- 75% free space blocks............0
50% -- 75% free space bytes.............0
75% -- 100% free space blocks...........12
75% -- 100% free space bytes............98304
Unused Blocks...........................0
Unused Bytes............................0
Total Blocks............................17
Total bytes.............................139264
PL/SQL procedure successfully completed
上面是沒降低HWM，如果載降低HWM，看看效果
SQL> alter table test3 shrink space cascade;
Table altered
SQL> exec show_space_t('TEST3','auto','table','Y');
Total Blocks............................24
Total Bytes.............................196608
Unused Blocks...........................0
Unused Bytes............................0
Last Used Ext FileId....................31
Last Used Ext BlockId...................481897
Last Used Block.........................8
*************************************************
0% -- 25% free space blocks.............1
0% -- 25% free space bytes..............8192
25% -- 50% free space blocks............2
25% -- 50% free space bytes.............16384
50% -- 75% free space blocks............0
50% -- 75% free space bytes.............0
75% -- 100% free space blocks...........0
75% -- 100% free space bytes............0
Unused Blocks...........................0
Unused Bytes............................0
Total Blocks............................17
Total bytes.............................139264
PL/SQL procedure successfully completed
看來用shrink space整理碎片不徹底，再來看看move的方式
SQL> alter table test3 move;
Table altered
SQL> exec show_space_t('TEST3','auto','table','Y');
Total Blocks............................32
Total Bytes.............................262144
Unused Blocks...........................6
Unused Bytes............................49152
Last Used Ext FileId....................31
Last Used Ext BlockId...................485081
Last Used Block.........................2
*************************************************
0% -- 25% free space blocks.............0
0% -- 25% free space bytes..............0
25% -- 50% free space blocks............0
25% -- 50% free space bytes.............0
50% -- 75% free space blocks............0
50% -- 75% free space bytes.............0
75% -- 100% free space blocks...........0
75% -- 100% free space bytes............0
Unused Blocks...........................0
Unused Bytes............................0
Total Blocks............................22
Total bytes.............................180224
PL/SQL procedure successfully completed
效果很明顯，整理的很徹底
測試結論：
雖然alter table move和shrink space，都是通過物理調整rowid來整理碎片的，但shrink space整理的不徹底，他好像不是重組，而是儘可能的合併，隨意會殘留一些block無法整理
注意：
1.再用alter table table_name move時，表相關的索引會失效，所以之後還要執行 alter index index_name rebuild online; 最後重新編譯資料庫所有失效的物件
2. 在用alter table table_name shrink space cascade時，他相當於alter table table_name move和alter index index_name rebuild online. 所以只要編譯資料庫失效的物件就可以
alter table move和shrink space除了碎片整理的效果有時不一樣外，還有什麼其他的不同呢
1. Move會移動高水位，但不會釋放申請的空間，是在高水位以下(below HWM)的操作。
2. shrink space 同樣會移動高水位，但也會釋放申請的空間，是在高水位上下(below and above HWM)都有的操作。
下面通過實驗來驗證
SQL> drop table test3;
Table dropped
SQL>
SQL> create table test3 as
2 select rownum id,
3 dbms_random.string('a', round(dbms_random.value(0,10))) col1,
4 trunc(sysdate) - dbms_random.value(1, 365*2) col2
5 from dual connect by rownum<=10000;
Table created
SQL> analyze table test3 compute statistics;
Table analyzed
SQL> col segment_name for a10;
SQL> select us.segment_name,us.extents,us.blocks from user_segments us where us.segment_name=upper('test3');
SEGMENT_NA EXTENTS BLOCKS
---------- ---------- ----------
TEST3 5 40
SQL> col table_name for a10;
SQL> select table_name,blocks,t.empty_blocks from user_tables t where t.table_name=upper('test3');
TABLE_NAME BLOCKS EMPTY_BLOCKS
---------- ---------- ------------
TEST3 37 3
SQL>
從以上查詢可以看出共分了5個extents，使用了37個blocks，這37也就是test3的HWM
SQL> delete from test3 where rownum<=5000;
5000 rows deleted
SQL> commit;
Commit complete
SQL> analyze table test3 compute statistics;
Table analyzed
SQL> col segment_name for a10;
SQL> select us.segment_name,us.extents,us.blocks from user_segments us where us.segment_name=upper('test3');
SEGMENT_NA EXTENTS BLOCKS
---------- ---------- ----------
TEST3 5 40
SQL> col table_name for a10;
SQL> select table_name,blocks,t.empty_blocks from user_tables t where t.table_name=upper('test3');
TABLE_NAME BLOCKS EMPTY_BLOCKS
---------- ---------- ------------
TEST3 37 3
SQL> select count(distinct dbms_rowid.rowid_block_number(rowid)) used_blocks from test3;
USED_BLOCKS
-----------
17
我們從查詢中可以發現test3的HWM沒有變換還是37blocks，tests總共空間為40blocks。經過刪除後test3實際用的塊是17個
下面我們用move降低下HWM
SQL> alter table test3 move;
Table altered
SQL> col segment_name for a10;
SQL> select us.segment_name,us.extents,us.blocks from user_segments us where us.segment_name=upper('test3');
SEGMENT_NA EXTENTS BLOCKS
---------- ---------- ----------
TEST3 3 24
SQL> col table_name for a10;
SQL> select table_name,blocks,t.empty_blocks from user_tables t where t.table_name=upper('test3');
TABLE_NAME BLOCKS EMPTY_BLOCKS
---------- ---------- ------------
TEST3 37 3
user_tables裡的資料沒有變化，哦，原來是忘記analyze了，從這裡也可以看出user_segments是oracle自動維護的。
SQL> analyze table test3 compute statistics;
Table analyzed
SQL> select us.segment_name,us.extents,us.blocks from user_segments us where us.segment_name=upper('test3');
SEGMENT_NA EXTENTS BLOCKS
---------- ---------- ----------
TEST3 3 24
SQL> select table_name,blocks,t.empty_blocks from user_tables t where t.table_name=upper('test3');
TABLE_NAME BLOCKS EMPTY_BLOCKS
---------- ---------- ------------
TEST3 20 4
SQL>
現在再來看hwm變為20了，已經降下來了啊，空間也收縮了，從40blocks降到24blocks（注意收縮到initial指定值）。
但shrink space就收縮到儲存資料的最小值，下面測試說明
建立測試表：
SQL> create table test5 (id number) storage (initial 1m next 1m);
Table created
初始化資料
SQL>
SQL> begin
2 for i in 1..100000 loop
3 insert into test5 values(i);
4 end loop;
5 end;
6 /
PL/SQL procedure successfully completed
SQL> analyze table test5 compute statistics;
Table analyzed
SQL> select SEGMENT_NAME,EXTENTS,BLOCKS,INITIAL_EXTENT/1024/1024 init from user_segments where SEGMENT_NAME='TEST5';
SEGMENT_NA EXTENTS BLOCKS INIT
---------- ---------- ---------- ----------
TEST5 17 256 1
SQL> select TABLE_NAME,BLOCKS,EMPTY_BLOCKS from user_tables where table_name='TEST5';
TABLE_NAME BLOCKS EMPTY_BLOCKS
---------- ---------- ------------
TEST5 180 76
可以從查詢資料看出，test5初始化1m即128個blocks，但資料比較多，所以又按next引數要求擴充套件了1m空間，擴充套件了17個extents。
這裡的test5總空間大小為256個blocks，使用空間為180blocks，HWM也是180blocks
SQL> delete from test5 where rownum<=50000;
50000 rows deleted
SQL> analyze table test5 compute statistics;
Table analyzed
SQL> select SEGMENT_NAME,EXTENTS,BLOCKS,INITIAL_EXTENT/1024/1024 init from user_segments where SEGMENT_NAME='TEST5';
SEGMENT_NA EXTENTS BLOCKS INIT
---------- ---------- ---------- ----------
TEST5 17 256 1
SQL> select TABLE_NAME,BLOCKS,EMPTY_BLOCKS from user_tables where table_name='TEST5';
TABLE_NAME BLOCKS EMPTY_BLOCKS
---------- ---------- ------------
TEST5 180 76
整理碎片，降低HWM
SQL> alter table test5 move;
Table altered
SQL> analyze table test5 compute statistics;
Table analyzed
SQL> select SEGMENT_NAME,EXTENTS,BLOCKS,INITIAL_EXTENT/1024/1024 init from user_segments where SEGMENT_NAME='TEST5';
SEGMENT_NA EXTENTS BLOCKS INIT
---------- ---------- ---------- ----------
TEST5 16 128 1
SQL> select TABLE_NAME,BLOCKS,EMPTY_BLOCKS from user_tables where table_name='TEST5';
TABLE_NAME BLOCKS EMPTY_BLOCKS
---------- ---------- ------------
TEST5 85 43
從上面的查詢資料可以看出HWM已經從180降低到85，test5總大小從256blocks收縮到128個blocks（initial指定大小）。
下面看看用shrink space收縮空間的情況
SQL> alter table test5 enable row movement;
Table altered
SQL> alter table test5 shrink space;
Table altered
SQL> select SEGMENT_NAME,EXTENTS,BLOCKS,INITIAL_EXTENT/1024/1024 init from user_segments where SEGMENT_NAME='TEST5';
SEGMENT_NA EXTENTS BLOCKS INIT
---------- ---------- ---------- ----------
TEST5 11 88 1
SQL> select TABLE_NAME,BLOCKS,EMPTY_BLOCKS from user_tables where table_name='TEST5';
TABLE_NAME BLOCKS EMPTY_BLOCKS
---------- ---------- ------------
TEST5 85 43
SQL>
從上面的資料可以看到test5進一步從128個blocks降低到88個blocks
結論：
shrink space收縮到資料儲存的最小值，alter table move（不帶引數）收縮到initial指定值，也可以用alter table test5 move storage(initial 500k)指定收縮的大小，這樣可以達到shrink space效果
經過以上測試，得出的兩個結論，到底用哪一個命令來整理碎片，消除行遷移呢？這就要根據實際業務需要，如果你只是收縮空間，資料增長很慢，那用shrink可以但是如果資料增長很快的話，用move就比較合適，避免再重新分配空間啊
備註：
在10g之後，整理碎片消除行遷移的新增功能shrink space
alter table
compact:這個引數當系統的負載比較大時可以用，不降低HWM。如果系統負載較低時，直接用alter table table_name shrink space就一步到位了
cascade：這個引數是在shrink table的時候自動級聯索引，相當於rebulid index。
普通表：
shrink必須開啟行遷移功能。
alter table table_name enable row movement ;
保持HWM，相當於把塊中資料打結實了
alter table table_name shrink space compact;
回縮表與降低HWM
alter table table_name shrink space;
回縮表與相關索引，降低HWM
alter table table_name shrink space cascade;
回縮索引與降低HWM
alter index index_name shrink space
雖然在10g中可以用shrink ，但也有些限制：
1). 對cluster，cluster table，或具有Long，lob型別列的物件 不起作用。
2). 不支援具有function-based indexes 或 bitmap join indexes的表
3). 不支援mapping 表或index-organized表。
4). 不支援compressed 表
來自 “ ITPUB部落格 ” ，連結：http://blog.itpub.net/24104518/viewspace-764904/，如需轉載，請註明出處，否則將追究法律責任。
相關文章
- alter table move跟shrink space的區別
- ALTER TABLE MOVE | SHRINK SPACE區別
- alter table move 和 alter table shrink space的區別
- ALTER TABLE MOVE和SHRINK SPACE區別
- alter table move 與shrink space的區別
- Oracle 11g alter table move與shrink spaceOracle
- table move 與 shrink 的區別
- [Oracle] Shrink space & Table move比較Oracle
- oracle 10g__alter table shrink space compactOracle 10g
- alter table table_name move ; 在自身表空間move是如何操作的？
- Oracle 10g Shrink Table - Shrink Space 收縮空間Oracle 10g
- oracle10g_alter table shrink space_compact_cascade回收空間測試(一)Oracle
- alter table列管理的一些區別
- 測試alter table shrink space compact cascade及學習user_tables相關列的含義
- 表、索引遷移表空間alter table move索引
- v$lock之alter table drop column與alter table set unused column區別系列五
- 【轉】Oracle：MOVE與SHRINK命令相比較Oracle
- Oracle IZ0-053 Q277(Table shrink space)Oracle
- How to adjust the high watermark in ORACLE 10g – ALTER TABLE SHRINKOracle 10g
- shrink space的最佳實踐
- 測試alter table storage及dbms_space_admin包
- oracle shrink tableOracle
- alter database和alter system和alter session的區別DatabaseSession
- [重慶思莊每日技術分享]-在為表新增了列後執行ALTER TABLE SHRINK SPACE 提示ORA-8102
- [轉]alter system switch logfile和alter system archive log current的區別Hive
- alter system events與alter system event的區別
- Oracle中shrink space命令詳解Oracle
- ALTER DATABASE 與 ALTER TABLESPACE OFFLINE的區別Database
- Oracle中shrink space命令詳解[轉]－－還示測試Oracle
- Alter table for ORACLEOracle
- 【轉】dbms_stats.gather_table_stats與analyze table 的區別
- mysql的ALTER TABLE命令MySql
- Oracle move和shrink釋放高水位空間Oracle
- 忍不住問下alter system 和alter database的區別Database
- drop table和truncate table的區別
- dbms_stats.gather_table_stats與analyze table 的區別[轉貼]
- oracle10g shrink space 降低HWMOracle
- 7 、shrink table and its dependent segments