行鏈接和行遷移案例

兩則之間的區(qū)別：

???????????? 行鏈接是指一個(gè)行存儲(chǔ)在多個(gè)塊中的情況，因?yàn)橐粋€(gè)該行的長(zhǎng)度超過(guò)了一個(gè)塊的可用空間大小。（Insert）
???????????? 行遷移是指一個(gè)數(shù)據(jù)行不適合放入當(dāng)前塊而被重新定位到另一個(gè)塊，但在原始?jí)K中保留一個(gè)指針，原始?jí)K中的指針是必需的，因?yàn)樗饕膔owid項(xiàng)仍然指向原始位置.（update）?????????

??????????? 行鏈接通常與行的長(zhǎng)度和oracle數(shù)據(jù)庫(kù)塊中的大小有關(guān)，而行遷移通常是當(dāng)一個(gè)更新操作的長(zhǎng)度增加且又要保持該行在同一塊中，而該塊又缺少可用空間時(shí)產(chǎn)生的問(wèn)題，oracle在決定行連接之前先試圖進(jìn)行行遷移。

行連接

1、建立表

04:50:36 SQL> show user;
USER is "SCOTT"
04:50:39 SQL>? create table t_row_chaining (x char(2000), y char(2000), z char(2000), q char(2000));

Table created.

04:50:51 SQL> insert into t_row_chaining values('x','y','z','q');

1 row created.

04:51:05 SQL> commit;

Commit complete.

2、執(zhí)行腳本，建立行鏈接和行遷移視圖

04:51:10 SQL> @?/rdbms/admin/utlchain.sql

Table created.

04:51:32 SQL> select * from tab;

TNAME????????????????????????? TABTYPE? CLUSTERID
------------------------------ ------- ----------
DEPT?????????????????????????? TABLE
EMP??????????????????????????? TABLE
BONUS????????????????????????? TABLE
SALGRADE?????????????????????? TABLE
EMP1?????????????????????????? TABLE
EMP5?????????????????????????? TABLE
SALES????????????????????????? TABLE
SYS_IOT_OVER_12021???????????? TABLE
SALES_INFO???????????????????? TABLE
DEPT_EMP_CLU?????????????????? CLUSTER
DEPARTMENT???????????????????? TABLE??????????? 1
EMPLOYEE?????????????????????? TABLE??????????? 2
TMP1?????????????????????????? TABLE
TMP2?????????????????????????? TABLE
T_ROW_CHAINING???????????????? TABLE
CHAINED_ROWS?????????????????? TABLE

16 rows selected.

04:51:40 SQL> desc CHAINED_ROWS
?Name??????????????????????????????????????????????????????????????????????????????? Null???? Type
?----------------------------------------------------------------------------------- -------- --------------------------------------------------------
?OWNER_NAME?????????????????????????????????????????????????????????????????????????????????? VARCHAR2(30)
?TABLE_NAME?????????????????????????????????????????????????????????????????????????????????? VARCHAR2(30)
?CLUSTER_NAME???????????????????????????????????????????????????????????????????????????????? VARCHAR2(30)
?PARTITION_NAME?????????????????????????????????????????????????????????????????????????????? VARCHAR2(30)
?SUBPARTITION_NAME??????????????????????????????????????????????????????????????????????????? VARCHAR2(30)
?HEAD_ROWID?????????????????????????????????????????????????????????????????????????????????? ROWID
?ANALYZE_TIMESTAMP???
?
?3、分析表的行鏈接信息???????????????????????????????????????????????????????????????????????? DATE

04:51:50 SQL> ANALYZE TABLE t_row_chaining LIST CHAINED ROWS;

Table analyzed.

04:52:09 SQL>

04:52:09 SQL> select OWNER_NAME,TABLE_NAME,HEAD_ROWID,ANALYZE_TIMESTAMP from chained_rows;

OWNER_NAME???????????????????? TABLE_NAME???????????????????? HEAD_ROWID???????? ANALYZE_T
------------------------------ ------------------------------ ------------------ ---------
SCOTT????????????????????????? T_ROW_CHAINING???????????????? AAAC9KAAEAAAAB3AAA 25-MAR-12

-------由于插入的記錄長(zhǎng)度超過(guò)了block的free space ，發(fā)生行鏈接

行遷移

1、建立表

04:53:06 SQL>? create table t_row_migrating (x varchar2(2000), y varchar2(2000), z varchar2(2000), q varchar2(2000));

Table created.

04:54:01 SQL> insert into t_row_migrating values ('x','y','z','q');

1 row created.

04:54:12 SQL> commit;

Commit complete.

04:54:20 SQL> delete chained_rows;

1 row deleted.

04:54:42 SQL> ANALYZE TABLE t_row_migrating LIST CHAINED ROWS;

Table analyzed.

04:54:54 SQL>? select OWNER_NAME,TABLE_NAME,HEAD_ROWID,ANALYZE_TIMESTAMP from chained_rows;

no rows selected

2、對(duì)表做update 實(shí)驗(yàn)

04:55:03 SQL>? update t_row_migrating set (x,y,z,q)=(select * from t_row_chaining);

1 row updated.

04:57:17 SQL> commit;

Commit complete.

04:57:33 SQL> ANALYZE TABLE t_row_migrating LIST CHAINED ROWS;

Table analyzed.

04:57:41 SQL>? select OWNER_NAME,TABLE_NAME,HEAD_ROWID,ANALYZE_TIMESTAMP from chained_rows;

OWNER_NAME???????????????????? TABLE_NAME???????????????????? HEAD_ROWID???????? ANALYZE_T
------------------------------ ------------------------------ ------------------ ---------
SCOTT????????????????????????? T_ROW_MIGRATING??????????????? AAAC9MAAEAAAACGAAA 25-MAR-12

04:57:53 SQL>

no rows selected

04:55:03 SQL>

----------更新后的記錄變長(zhǎng)，在block free space 無(wú)法存放，發(fā)生行遷移

本文轉(zhuǎn)自 客居天涯 51CTO博客，原文鏈接：http://blog.51cto.com/tiany/816715，如需轉(zhuǎn)載請(qǐng)自行聯(lián)系原作者

總結(jié)

以上是生活随笔為你收集整理的行链接和行迁移案例的全部?jī)?nèi)容，希望文章能夠幫你解決所遇到的問(wèn)題。

如果覺(jué)得生活随笔網(wǎng)站內(nèi)容還不錯(cuò)，歡迎將生活随笔推薦給好友。