Oracle表碎片整理操作步驟詳解

2020-07-26 14:22:19

字體：大中小

供稿：網(wǎng)友

高水位線(xiàn)（HWL）下的許多數(shù)據(jù)塊都是無(wú)數(shù)據(jù)的，但全表掃描的時(shí)候要掃描到高水位線(xiàn)的數(shù)據(jù)塊，也就是說(shuō)oracle要做許多的無(wú)用功！因此oracle提供了shrink space碎片整理功能。對(duì)于索引，可以采取rebuild online的方式進(jìn)行碎片整理，一般來(lái)說(shuō)，經(jīng)常進(jìn)行DML操作的對(duì)象DBA要定期進(jìn)行維護(hù)，同時(shí)注意要及時(shí)更新統(tǒng)計(jì)信息！

一：準(zhǔn)備測(cè)試數(shù)據(jù)，使用HR用戶(hù)，創(chuàng)建T1表，插入約30W的數(shù)據(jù)，并根據(jù)object_id創(chuàng)建普通索引，表占存儲(chǔ)空間34M

復(fù)制代碼代碼如下:

SQL> conn /as sysdba 
已連接。 
SQL> select default_tablespace from dba_users where username='HR'; 

DEFAULT_TABLESPACE 
------------------------------------------------------------ 
USERS 

SQL> conn hr/hr 
已連接。 

SQL> insert into t1 select * from t1; 
已創(chuàng)建 74812 行。 

SQL> insert into t1 select * from t1; 
已創(chuàng)建 149624 行。 

SQL> commit; 
提交完成。 

SQL> create index idx_t1_id on t1(object_id); 
索引已創(chuàng)建。 

SQL> exec dbms_stats.gather_table_stats('HR','T1',CASCADE=>TRUE); 
PL/SQL 過(guò)程已成功完成。 

SQL> select count(1) from t1; 

  COUNT(1) 
---------- 
    299248 

SQL> select sum(bytes)/1024/1024 from dba_segments where segment_name='T1'; 
SUM(BYTES)/1024/1024 
-------------------- 
             34.0625 

SQL> select sum(bytes)/1024/1024 from dba_segments where segment_name='IDX_T1_ID'; 
SUM(BYTES)/1024/1024 
-------------------- 
                   6 

二：估算表在高水位線(xiàn)下還有多少空間可用，這個(gè)值應(yīng)當(dāng)越低越好，表使用率越接近高水位線(xiàn)，全表掃描所做的無(wú)用功也就越少!

DBMS_STATS包無(wú)法獲取EMPTY_BLOCKS統(tǒng)計(jì)信息，所以需要用analyze命令再收集一次統(tǒng)計(jì)信息

復(fù)制代碼代碼如下:

SQL> SELECT blocks, empty_blocks, num_rows FROM user_tables WHERE table_name ='T1'; 

    BLOCKS EMPTY_BLOCKS   NUM_ROWS 
---------- ------------ ---------- 
      4302            0     299248 

SQL> analyze table t1 compute statistics; 
表已分析。 

SQL> SELECT blocks, empty_blocks, num_rows FROM user_tables WHERE table_name ='T1'; 

    BLOCKS EMPTY_BLOCKS   NUM_ROWS 
---------- ------------ ---------- 
      4302           50     299248 

SQL> col table_name for a20 
SQL> SELECT TABLE_NAME, 
  2         (BLOCKS * 8192 / 1024 / 1024) - 
  3         (NUM_ROWS * AVG_ROW_LEN / 1024 / 1024) "Data lower than HWM in MB" 
  4    FROM USER_TABLES 
  5   WHERE table_name = 'T1'; 

TABLE_NAME           Data lower than HWM in MB 
-------------------- ------------------------- 
T1                                  5.07086182 

三: 查看執(zhí)行計(jì)劃，全表掃描大概需要消耗CPU 1175

復(fù)制代碼代碼如下:

SQL> explain plan for select * from t1; 
已解釋。 

SQL> select * from table(dbms_xplan.display); 

PLAN_TABLE_OUTPUT 
-------------------------------------------------------------------------------- 
Plan hash value: 3617692013 
-------------------------------------------------------------------------- 
| Id  | Operation         | Name | Rows  | Bytes | Cost (%CPU)| Time     | 
-------------------------------------------------------------------------- 
|   0 | SELECT STATEMENT  |      |   299K|    28M|  1175   (1)| 00:00:15 | 
|   1 |  TABLE ACCESS FULL| T1   |   299K|    28M|  1175   (1)| 00:00:15 | 
-------------------------------------------------------------------------- 

四：刪除大部分?jǐn)?shù)據(jù)，收集統(tǒng)計(jì)信息，全表掃描依然需要消耗CPU 1168

復(fù)制代碼代碼如下:

SQL> delete from t1 where object_id>100; 
已刪除298852行。 

SQL> commit; 
提交完成。 

SQL> select count(*) from t1; 

  COUNT(*) 
---------- 
       396 

SQL>  exec dbms_stats.gather_table_stats('HR','T1',CASCADE=>TRUE); 
PL/SQL 過(guò)程已成功完成。 

SQL> analyze table t1 compute statistics; 
表已分析。 

SQL> SELECT blocks, empty_blocks, num_rows FROM user_tables WHERE table_name ='T1'; 

    BLOCKS EMPTY_BLOCKS   NUM_ROWS 
---------- ------------ ---------- 
      4302           50        396 

 
SQL> explain plan for select * from t1; 
已解釋。 

SQL> select * from table(dbms_xplan.display); 

PLAN_TABLE_OUTPUT 
------------------------------------------------------------------------------ 
Plan hash value: 3617692013 
-------------------------------------------------------------------------- 
| Id  | Operation         | Name | Rows  | Bytes | Cost (%CPU)| Time     | 
-------------------------------------------------------------------------- 
|   0 | SELECT STATEMENT  |      |   396 | 29700 |  1168   (1)| 00:00:15 | 
|   1 |  TABLE ACCESS FULL| T1   |   396 | 29700 |  1168   (1)| 00:00:15 | 
-------------------------------------------------------------------------- 

五：估算表在高水位線(xiàn)下還有多少空間是無(wú)數(shù)據(jù)的，但在全表掃描時(shí)又需要做無(wú)用功的數(shù)據(jù)

復(fù)制代碼代碼如下:

SQL> SELECT TABLE_NAME, 
  2         (BLOCKS * 8192 / 1024 / 1024) - 
  3         (NUM_ROWS * AVG_ROW_LEN / 1024 / 1024) "Data lower than HWM in MB" 
  4    FROM USER_TABLES 
  5   WHERE table_name = 'T1'; 

TABLE_NAME           Data lower than HWM in MB 
-------------------- ------------------------- 
T1                                  33.5791626 

六：對(duì)表進(jìn)行碎片整理，重新收集統(tǒng)計(jì)信息

復(fù)制代碼代碼如下:

SQL> alter table t1 enable row movement; 
表已更改。 

SQL> alter table t1 shrink space cascade; 
表已更改。 

SQL> select sum(bytes)/1024/1024 from dba_segments where segment_name='T1'; 

SUM(BYTES)/1024/1024 
-------------------- 
                .125 

SQL> select sum(bytes)/1024/1024 from dba_segments where segment_name='IDX_T1_ID 
'; 

SUM(BYTES)/1024/1024 
-------------------- 
               .0625 

SQL> SELECT TABLE_NAME, 
  2         (BLOCKS * 8192 / 1024 / 1024) - 
  3         (NUM_ROWS * AVG_ROW_LEN / 1024 / 1024) "Data lower than HWM in MB" 
  4    FROM USER_TABLES 
  5   WHERE table_name = 'T1'; 

TABLE_NAME           Data lower than HWM in MB 
-------------------- ------------------------- 
T1                                  33.5791626 

SQL> exec dbms_stats.gather_table_stats('HR','T1',CASCADE=>TRUE); 
PL/SQL 過(guò)程已成功完成。 

這個(gè)時(shí)候，只剩下0.1M的無(wú)用功了，執(zhí)行計(jì)劃中，全表掃描也只需要消耗CPU 3 
SQL> SELECT TABLE_NAME, 
  2         (BLOCKS * 8192 / 1024 / 1024) - 
  3         (NUM_ROWS * AVG_ROW_LEN / 1024 / 1024) "Data lower than HWM in MB" 
  4    FROM USER_TABLES 
  5   WHERE table_name = 'T1'; 

TABLE_NAME           Data lower than HWM in MB 
-------------------- ------------------------- 
T1                                  .010738373 

 
SQL> select * from table(dbms_xplan.display); 

PLAN_TABLE_OUTPUT 
-------------------------------------------------------------------------------- 
Plan hash value: 3617692013 
-------------------------------------------------------------------------- 
| Id  | Operation         | Name | Rows  | Bytes | Cost (%CPU)| Time     | 
-------------------------------------------------------------------------- 
|   0 | SELECT STATEMENT  |      |   396 | 29700 |     3   (0)| 00:00:01 | 
|   1 |  TABLE ACCESS FULL| T1   |   396 | 29700 |     3   (0)| 00:00:01 | 
-------------------------------------------------------------------------- 

總共只有5個(gè)塊，空塊卻有50個(gè)，明顯empty_blocks信息過(guò)期 
SQL> select blocks,empty_blocks,num_rows from user_tables where table_name='T1'; 

    BLOCKS EMPTY_BLOCKS   NUM_ROWS 
---------- ------------ ---------- 
         5           50        396 

SQL> analyze table t1 compute statistics; 
表已分析。 

SQL> select blocks,empty_blocks,num_rows from user_tables where table_name='T1'; 

 
    BLOCKS EMPTY_BLOCKS   NUM_ROWS 
---------- ------------ ---------- 
         5            3        396 