一、参考[学习DBA 范计杰]
https://www.modb.pro/db/40753 http://www.360doc.com/content/21/0121/05/70704971_958071999.shtml Purging RECYCLEBIN Running Fosrever (Doc ID 2284986.1)
二、概述
某客户的数据库一条insert语句某段时间突然变慢，平均单次执行逻辑读暴增至20万，最终定位为insert时因可用空间不充足递归清理recyclebin中的对象导致，
递归delete from RECYCLEBIN$，因没有合适的索引走full scan，正常情况下也没啥影响，但该客户的这套库SYS.RECYCLEBIN$竟然达到700M.最终导致了该问题。
因为是学习案例，并没有接触这个案例，需要重新梳理： 1.客户是执行Insert语句突然变慢的问题; 2.后续定位是递归SQL中,存在delete from RECYCLEBIN$逻辑读很高,发现问题<最终创建合适索引,并且对基表收集统计信息后解决问题！ 问题： 1.什么情况下会导致这个问题？ 2.如何最快速度判断是这个问题？ 3.如何避免出现这个问题？ 4.正常drop table, drop table purge,这个操作与RECYCLEBIN$基表的数据有什么关联？ 如下测试围绕着上述问题进行测试,并进行小结。
三、测试
3.1 drop table, drop table purge,这个操作与RECYCLEBIN$基表的数据有什么关联？
1）清空回收站！ SQL> select count(*) from RECYCLEBIN$; COUNT(*) ---------- 14 SQL> purge dba_recyclebin; DBA Recyclebin purged. SQL> select count(*) from RECYCLEBIN$; COUNT(*) ---------- 0 可以发现，清空回收站后，基表RECYCLEBIN$记录被删除。 2）drop table xx purge; SQL> conn scott/tiger SQL> drop table a purge; Table dropped. SQL> show recyclebin; 无记录 3）drop table xx 那么在基表会有多少条记录呢？ 一条？ 两条？？？ SQL> create table a as select * from dba_objects; SQL> select count(extent_id) from user_extents where segment_name='A'; COUNT(EXTENT_ID) ---------------- 25 SQL> drop table a; SQL> select count(*) from RECYCLEBIN$; COUNT(*) ---------- 1 SQL> select * from RECYCLEBIN$; OBJ# OWNER# ORIGINAL_NAME OPERATION TYPE# ---------- ---------- -------------------------------- ---------- ---------- TS# FILE# BLOCK# DROPTIME DROPSCN ---------- ---------- ---------- --------- ---------- PARTITION_NAME FLAGS RELATED BO PURGEOBJ -------------------------------- ---------- ---------- ---------- ---------- BASE_TS# BASE_OWNER# SPACE CON# SPARE1 SPARE2 SPARE3 ---------- ----------- ---------- ---------- ---------- ---------- ---------- 90660 83 A 0 1 4 4 4154 17-JAN-21 35784550 30 90660 90660 90660 1280 0 SQL> show recyclebin; ORIGINAL NAME RECYCLEBIN NAME OBJECT TYPE DROP TIME ---------------- ------------------------------ ------------ ------------------- A BIN$uQgoIhFLEGngU0U8qMB9Sg==$0 TABLE 2021-01-17:00:41:12
？ 也就是说多个extents 但是基表只占用1条记录？
4）测试使用一个分区表！ 进行drop 测试
SQL> select SEGMENT_NAME,PARTITION_NAME from user_segments where segment_name='RANGE_PART_TAB1' order by 2;
SEGMENT_NAME PARTITION_NAME
------------------------------ ------------------------------
RANGE_PART_TAB1 P1
······
RANGE_PART_TAB1 P_MAX
13 rows selected.
SQL> drop table RANGE_PART_TAB1;
SQL> show recyclebin;
ORIGINAL NAME RECYCLEBIN NAME OBJECT TYPE DROP TIME
---------------- ------------------------------ ------------ -------------------
A BIN$uQgoIhFLEGngU0U8qMB9Sg==$0 TABLE 2021-01-17:00:41:12
RANGE_PART_TAB1 BIN$uQgoIhFMEGngU0U8qMB9Sg==$0 TABLE 2021-01-17:02:49:25
SQL> select count(*) from RECYCLEBIN$;
COUNT(*)
----------
15
也就是说非分区表,drop table 1条记录！ 分区表，一个segment drop table 将占用多少记录进行存储至基表RECYCLEBIN$中
所以也就是说，当数据库经常出现drop table的操作或者删除大量分区表，并且没有加上purge,也没有定期清空回收站的情况下，容易造成回收站基表的增大！
3.2 模拟出现insert 慢，源头是delete 回收站基表SQL慢的情况
将基表Segment大小变大！ 目的是后续Insert时,申请空间不足，需要delete基表SQL执行很慢，从而影响insert 速度！
SQL> select count(*) from recyclebin$;
COUNT(*)
----------
15
SQL> select bytes/1024/1024 from user_segments where segment_name='RECYCLEBIN$';
BYTES/1024/1024
---------------
.0625
SQL>insert into sys.RECYCLEBIN$ select t.* from sys.RECYCLEBIN$ t connect by level<2000000
*
ERROR at line 1:
ORA-01013: user requested cancel of current operation
SQL> select count(*) from recyclebin$;
COUNT(*)
----------
15
select segment_name,segment_type,bytes/1024/1024 from user_segments where segment_name like 'RECYCLEBIN%'；
SEGMENT_NAME SEGMENT_TYPE BYTES/1024/1024
------------------------- ------------------ ---------------
RECYCLEBIN$ TABLE 480
RECYCLEBIN$_OBJ INDEX 160
RECYCLEBIN$_TS INDEX 136
RECYCLEBIN$_OWNER INDEX 144
现在基表Segment达到了480Mbytes.
SQL> purge dba_recyclebin;
select segment_name,segment_type,bytes/1024/1024 from user_segments where segment_name like 'RECYCLEBIN%'
SEGMENT_NAME SEGMENT_TYPE BYTES/1024/1024
------------------------- ------------------ ---------------
RECYCLEBIN$ TABLE 480
RECYCLEBIN$_OBJ INDEX 160
RECYCLEBIN$_TS INDEX 136
RECYCLEBIN$_OWNER INDEX 144
基表没有变化！ 说明清空回收站并不会影响基表的大小！
制造一个很小的表空间
SQL> create tablespace testtbs datafile '/11.2.0.4/app/oracle/tt/testtbs.dbf' size 5m;
SQL> create table scott.tmp1 tablespace testtbs as select * from dba_objects where rownum<=10000;
SQL> SELECT OWNER,SEGMENT_NAME,SEGMENT_TYPE,TABLESPACE_NAME,BYTES/1024/1024,EXTENTS FROM DBA_SEGMENTS WHERE OWNER='SCOTT' AND SEGMENT_NAME='TMP1';
OWNER SEGMENT_NAME SEGMENT_TYPE TABLESPACE_NAME BYTES/1024/1024 EXTENTS
------------------------------ ------------------------- ------------------ ------------------------------ --------------- ----------
SCOTT TMP1 TABLE TESTTBS 2 17
SQL> createtable SCOTT.tmp2 tablespace testtbs asselect * from dba_objects where1=0;
droptable SCOTT.tmp1 ;
SQL> SELECT * FROM DBA_FREE_SPACE WHERE TABLESPACE_NAME='TESTTBS';
TABLESPACE_NAME FILE_ID BLOCK_ID BYTES BLOCKS RELATIVE_FNO
------------------------------ ------- ---------- ---------- ---------- ------------
TESTTBS 10 384 2097152 256 10
TESTTBS 10 128 65536 8 10
TESTTBS 10 136 65536 8 10
TESTTBS 10 144 65536 8 10
TESTTBS 10 152 65536 8 10
TESTTBS 10 160 65536 8 10
TESTTBS 10 168 65536 8 10
TESTTBS 10 176 65536 8 10
TESTTBS 10 184 65536 8 10
TESTTBS 10 192 65536 8 10
TESTTBS 10 200 65536 8 10
TESTTBS 10 208 65536 8 10
TESTTBS 10 216 65536 8 10
TESTTBS 10 224 65536 8 10
TESTTBS 10 232 65536 8 10
TESTTBS 10 240 65536 8 10
TESTTBS 10 248 65536 8 10
TESTTBS 10 256 1048576 128 10
18 rows selected.
SQL> SELECT SUM(BYTES)/1024/1024 FROM DBA_FREE_SPACE WHERE TABLESPACE_NAME='TESTTBS';
SUM(BYTES)/1024/1024
--------------------
4
从dba_free_space block_id 248 ,连续8个blocks 248+7=255,连续下一个区的第一个block_id=256!
block_id 256 ,连续128个blocks
也就是说这个表空间文件头部占用了1Mbytes,第一次分配给tmp1 2Mbytes空间，drop table后标记为空闲！,其次是表空间数据文件创建后，剩余的2Mbytes连续未分配的空间block_id=384!
按照上述测试，插入1万条记录占用2Mbytes,那么如果一次插入1万条记录占用2Mbytes可以直接从free使用！
再次插入1万条记录，Oracle drop table的底层数据并未真正删除，还在回收站，因此需要内部执行SQL，其中涉及delete基表记录，SQL跑的很慢！
测试对比insert SQL执行效率
SQL> select count(*) from recyclebin$; COUNT(*) ---------- 1
SQL> oradebug setmypid Statement processed. SQL> oradebug event 10046 trace name context forever,level 12 Statement processed. SQL> oradebug tracefile_name /u01/app/oracle/diag/rdbms/tt/tt/trace/tt_ora_3996.trc
set timing on insert /*+gather_plan_statistics */into scott.tmp2 select * from dba_objects where rownum<=10000;
10000 rows created.
Elapsed: 00:00:00.28
set linesize 200 pagesize 999 select * from table(dbms_xplan.display_cursor(null,null,'advanced -PROJECTION -bytes iostats,last'));
Plan hash value: 310638681
------------------------------------------------------------------------------------------------------------------------------------------
| Id | Operation | Name | Starts | E-Rows | Cost (%CPU)| E-Time | A-Rows | A-Time | Buffers | Reads |
------------------------------------------------------------------------------------------------------------------------------------------
| 0 | INSERT STATEMENT | | 1 | | 303 (100)| | 0 |00:00:00.10 | 2580 | 15 |
| 1 | LOAD TABLE CONVENTIONAL | | 1 | | | | 0 |00:00:00.10 | 2580 | 15 |
|* 2 | COUNT STOPKEY | | 1 | | | | 10000 |00:00:00.07 | 124 | 0 |
| 3 | VIEW | DBA_OBJECTS | 1 | 86957 | 303 (1)| 00:00:04 | 10000 |00:00:00.07 | 124 | 0 |
| 4 | UNION-ALL | | 1 | | | | 10000 |00:00:00.07 | 124 | 0 |
|* 5 | TABLE ACCESS BY INDEX ROWID| SUM$ | 0 | 1 | 1 (0)| 00:00:01 | 0 |00:00:00.01 | 0 | 0 |
|* 6 | INDEX UNIQUE SCAN | I_SUM$_1 | 0 | 1 | 0 (0)| | 0 |00:00:00.01 | 0 | 0 |
| 7 | TABLE ACCESS BY INDEX ROWID| OBJ$ | 0 | 1 | 3 (0)| 00:00:01 | 0 |00:00:00.01 | 0 | 0 |
|* 8 | INDEX RANGE SCAN | I_OBJ1 | 0 | 1 | 2 (0)| 00:00:01 | 0 |00:00:00.01 | 0 | 0 |
|* 9 | FILTER | | 1 | | | | 10000 |00:00:00.04 | 124 | 0 |
|* 10 | HASH JOIN | | 1 | 86955 | 299 (2)| 00:00:04 | 10000 |00:00:00.03 | 124 | 0 |
| 11 | TABLE ACCESS FULL | USER$ | 1 | 99 | 3 (0)| 00:00:01 | 104 |00:00:00.01 | 6 | 0 |
|* 12 | HASH JOIN | | 1 | 86955 | 295 (1)| 00:00:04 | 10000 |00:00:00.02 | 118 | 0 |
| 13 | INDEX FULL SCAN | I_USER2 | 1 | 99 | 1 (0)| 00:00:01 | 104 |00:00:00.01 | 1 | 0 |
|* 14 | TABLE ACCESS FULL | OBJ$ | 1 | 86955 | 294 (1)| 00:00:04 | 10000 |00:00:00.01 | 117 | 0 |
| 15 | NESTED LOOPS | | 0 | 1 | 2 (0)| 00:00:01 | 0 |00:00:00.01 | 0 | 0 |
|* 16 | INDEX SKIP SCAN | I_USER2 | 0 | 1 | 1 (0)| 00:00:01 | 0 |00:00:00.01 | 0 | 0 |
|* 17 | INDEX RANGE SCAN | I_OBJ4 | 0 | 1 | 1 (0)| 00:00:01 | 0 |00:00:00.01 | 0 | 0 |
| 18 | NESTED LOOPS | | 0 | 2 | 4 (0)| 00:00:01 | 0 |00:00:00.01 | 0 | 0 |
| 19 | TABLE ACCESS FULL | LINK$ | 0 | 2 | 2 (0)| 00:00:01 | 0 |00:00:00.01 | 0 | 0 |
| 20 | TABLE ACCESS CLUSTER | USER$ | 0 | 1 | 1 (0)| 00:00:01 | 0 |00:00:00.01 | 0 | 0 |
|* 21 | INDEX UNIQUE SCAN | I_USER# | 0 | 1 | 0 (0)| | 0 |00:00:00.01 | 0 | 0 |
------------------------------------------------------------------------------------------------------------------------------------------
Predicate Information (identified by operation id):
---------------------------------------------------
2 - filter(ROWNUM<=10000)
5 - filter(BITAND("S"."XPFLAGS",8388608)=8388608)
6 - access("S"."OBJ#"=:B1)
8 - access("EO"."OBJ#"=:B1)
9 - filter((("O"."TYPE#"<>4 AND "O"."TYPE#"<>5 AND "O"."TYPE#"<>7 AND "O"."TYPE#"<>8 AND "O"."TYPE#"<>9 AND "O"."TYPE#"<>11
AND "O"."TYPE#"<>12 AND "O"."TYPE#"<>13 AND "O"."TYPE#"<>14 AND "O"."TYPE#"<>22 AND "O"."TYPE#"<>87 AND "O"."TYPE#"<>88) OR
BITAND("U"."SPARE1",16)=0 OR (INTERNAL_FUNCTION("O"."TYPE#") AND ((SYS_CONTEXT('userenv','current_edition_name')='ORA$BASE' AND
"U"."TYPE#"<>2) OR ("U"."TYPE#"=2 AND "U"."SPARE2"=TO_NUMBER(SYS_CONTEXT('userenv','current_edition_id'))) OR IS NOT NULL))))
10 - access("O"."SPARE3"="U"."USER#")
12 - access("O"."OWNER#"="U"."USER#")
14 - filter(("O"."TYPE#"<>10 AND "O"."NAME"<>'_NEXT_OBJECT' AND "O"."NAME"<>'_default_auditing_options_' AND "O"."LINKNAME" IS
NULL AND BITAND("O"."FLAGS",128)=0))
16 - access("U2"."TYPE#"=2 AND "U2"."SPARE2"=TO_NUMBER(SYS_CONTEXT('userenv','current_edition_id')))
filter(("U2"."TYPE#"=2 AND "U2"."SPARE2"=TO_NUMBER(SYS_CONTEXT('userenv','current_edition_id'))))
17 - access("O2"."DATAOBJ#"=:B1 AND "O2"."TYPE#"=88 AND "O2"."OWNER#"="U2"."USER#")
21 - access("L"."OWNER#"="U"."USER#")
再次Insert
SQL> insert /*+gather_plan_statistics */into scott.tmp2 select * from dba_objects where rownum<=10000; 10000 rows created. Elapsed: 00:00:00.63
可能从时间上，看起来没啥区别！ 数据量少，我们来看执行计划
Plan hash value: 310638681 ------------------------------------------------------------------------------------------------------------------------------------------ | Id | Operation | Name | Starts | E-Rows | Cost (%CPU)| E-Time | A-Rows | A-Time | Buffers | Reads | ------------------------------------------------------------------------------------------------------------------------------------------ | 0 | INSERT STATEMENT | | 1 | | 303 (100)| | 0 |00:00:00.63 | 123K| 81 | | 1 | LOAD TABLE CONVENTIONAL | | 1 | | | | 0 |00:00:00.63 | 123K| 81 | |* 2 | COUNT STOPKEY | | 1 | | | | 10000 |00:00:00.07 | 124 | 0 | | 3 | VIEW | DBA_OBJECTS | 1 | 86957 | 303 (1)| 00:00:04 | 10000 |00:00:00.07 | 124 | 0 | | 4 | UNION-ALL | | 1 | | | | 10000 |00:00:00.07 | 124 | 0 | |* 5 | TABLE ACCESS BY INDEX ROWID| SUM$ | 0 | 1 | 1 (0)| 00:00:01 | 0 |00:00:00.01 | 0 | 0 | |* 6 | INDEX UNIQUE SCAN | I_SUM$_1 | 0 | 1 | 0 (0)| | 0 |00:00:00.01 | 0 | 0 | | 7 | TABLE ACCESS BY INDEX ROWID| OBJ$ | 0 | 1 | 3 (0)| 00:00:01 | 0 |00:00:00.01 | 0 | 0 | |* 8 | INDEX RANGE SCAN | I_OBJ1 | 0 | 1 | 2 (0)| 00:00:01 | 0 |00:00:00.01 | 0 | 0 | |* 9 | FILTER | | 1 | | | | 10000 |00:00:00.04 | 124 | 0 | |* 10 | HASH JOIN | | 1 | 86955 | 299 (2)| 00:00:04 | 10000 |00:00:00.03 | 124 | 0 | | 11 | TABLE ACCESS FULL | USER$ | 1 | 99 | 3 (0)| 00:00:01 | 104 |00:00:00.01 | 6 | 0 | |* 12 | HASH JOIN | | 1 | 86955 | 295 (1)| 00:00:04 | 10000 |00:00:00.02 | 118 | 0 | | 13 | INDEX FULL SCAN | I_USER2 | 1 | 99 | 1 (0)| 00:00:01 | 104 |00:00:00.01 | 1 | 0 | |* 14 | TABLE ACCESS FULL | OBJ$ | 1 | 86955 | 294 (1)| 00:00:04 | 10000 |00:00:00.01 | 117 | 0 | | 15 | NESTED LOOPS | | 0 | 1 | 2 (0)| 00:00:01 | 0 |00:00:00.01 | 0 | 0 | |* 16 | INDEX SKIP SCAN | I_USER2 | 0 | 1 | 1 (0)| 00:00:01 | 0 |00:00:00.01 | 0 | 0 | |* 17 | INDEX RANGE SCAN | I_OBJ4 | 0 | 1 | 1 (0)| 00:00:01 | 0 |00:00:00.01 | 0 | 0 | | 18 | NESTED LOOPS | | 0 | 2 | 4 (0)| 00:00:01 | 0 |00:00:00.01 | 0 | 0 | | 19 | TABLE ACCESS FULL | LINK$ | 0 | 2 | 2 (0)| 00:00:01 | 0 |00:00:00.01 | 0 | 0 | | 20 | TABLE ACCESS CLUSTER | USER$ | 0 | 1 | 1 (0)| 00:00:01 | 0 |00:00:00.01 | 0 | 0 | |* 21 | INDEX UNIQUE SCAN | I_USER# | 0 | 1 | 0 (0)| | 0 |00:00:00.01 | 0 | 0 | ------------------------------------------------------------------------------------------------------------------------------------------
同样的执行计划？ 为什么Buffers 第一次Insert是2580，第二次123K ！！！
SQL>oradebug event 10046 trace name context off
$tkprof /u01/app/oracle/diag/rdbms/tt/tt/trace/tt_ora_3996.trc tk.txt
$ cat tk.txt |grep -A5 disk >disk.txt
通过人肉扫描可以发现query 60835较高的消耗
call count cpu elapsed disk query current rows ------- ------ -------- ---------- ---------- ---------- ---------- ---------- Parse 1 0.00 0.00 0 0 0 0 Execute 1 0.00 0.00 0 0 0 0 Fetch 2 0.16 0.16 0 60835 0 1 total 4 0.16 0.16 0 60835 0 1
$vi xx 定位这个区域
******************************************************************************** SQL ID: 0hhmdwwgxbw0r Plan Hash: 652194454 select obj#, type#, flags, related, bo, purgeobj, con# from RecycleBin$ where ts#=:1 and to_number(bitand(flags, 16)) = 16 order by dropscn call count cpu elapsed disk query current rows ------- ------ -------- ---------- ---------- ---------- ---------- ---------- Parse 1 0.00 0.00 0 0 0 0 Execute 1 0.00 0.00 0 0 0 0 Fetch 2 0.16 0.16 0 60835 0 1 ------- ------ -------- ---------- ---------- ---------- ---------- ---------- total 4 0.16 0.16 0 60835 0 1 Misses in library cache during parse: 1 Misses in library cache during execute: 1 Optimizer mode: CHOOSE Parsing user id: SYS (recursive depth: 1) Number of plan statistics captured: 1 Rows (1st) Rows (avg) Rows (max) Row Source Operation ---------- ---------- ---------- --------------------------------------------------- 1 1 1 SORT ORDER BY (cr=60835 pr=0 pw=0 time=163601 us cost=3 size=37 card=1) 1 1 1 TABLE ACCESS FULL RECYCLEBIN$ (cr=60835 pr=0 pw=0 time=163579 us cost=2 size=37 card=1) ********************************************************************************
drop table "SCOTT"."BIN$uQz/Q0JVEMfgU0U8qMDA7w==$0" purge
可以发现是先查询这个基表，随后drop table xx purge 清空回收站的记录！！！
******************************************************************************** SQL ID: b52m6vduutr8j Plan Hash: 716146596 delete from RecycleBin$ where bo=:1 call count cpu elapsed disk query current rows ------- ------ -------- ---------- ---------- ---------- ---------- ---------- Parse 1 0.00 0.00 0 0 0 0 Execute 1 0.13 0.13 0 60835 10 1 Fetch 0 0.00 0.00 0 0 0 0 ------- ------ -------- ---------- ---------- ---------- ---------- ---------- total 2 0.13 0.13 0 60835 10 1 Misses in library cache during parse: 1 Misses in library cache during execute: 1 Optimizer mode: CHOOSE Parsing user id: SYS (recursive depth: 1) Number of plan statistics captured: 1 Rows (1st) Rows (avg) Rows (max) Row Source Operation ---------- ---------- ---------- --------------------------------------------------- 0 0 0 DELETE RECYCLEBIN$ (cr=60835 pr=0 pw=0 time=136017 us) 1 1 1 TABLE ACCESS FULL RECYCLEBIN$ (cr=60835 pr=0 pw=0 time=135694 us cost=2 size=16 card=1) ********************************************************************************
在这个trace文件中，发现了两个调用RECYCLEBIN$的SQL，都是全表扫描，逻辑读消耗高！！！
四、问题处理
create index RecycleBin$_bo on RecycleBin$(bo);
exec dbms_stats.gather_table_stats(ownname=>'SYS',tabname=>'RecycleBin$', cascade=>TRUE);
select instance_number,sql_id,name, datatype_string,last_captured,value_string from dba_hist_sqlbind where sql_id='&sql'
and rownum<10 order by LAST_CAPTURED,POSITION; INSTANCE_NUMBER SQL_ID NAME DATATYPE_STRI LAST_CAPTURED VALUE_STRING --------------- --------------- ---------- ------------- ------------------------ ------------------------------ 1 0hhmdwwgxbw0r :1 NUMBER 17-JAN-21 12
1 b52m6vduutr8j :1 NUMBER 17-JAN-21 90695
测试！
EXPLAIN PLAN FOR select obj#, type#, flags, related, bo, purgeobj, con#
from RecycleBin$ where ts#=12 and to_number(bitand(flags, 16)) = 16 order by dropscn;
SQL> SELECT * FROM TABLE(DBMS_XPLAN.DISPLAY); PLAN_TABLE_OUTPUT
-------------------------------------------------------
Plan hash value: 885754405
--------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time |
-------------------------------------------------------------------------
| 0 | SELECT STATEMENT | | 1 | 117 | 3 (34)| 00:00:01 |
| 1 | SORT ORDER BY | | 1 | 117 | 3 (34)| 00:00:01 |
|* 2 | TABLE ACCESS BY INDEX ROWID| RECYCLEBIN$ | 1 | 117 | 2 (0)| 00:00:01 |
|* 3 | INDEX RANGE SCAN | RECYCLEBIN$_TS | 1 | | 2 (0)| 00:00:01 |
---------- Predicate Information (identified by operation id):
---------------------------------------------------
2 - filter(TO_NUMBER(TO_CHAR(BITAND("FLAGS",16)))=16)
3 - access("TS#"=12)
??? 有索引，之前没走？？？ 原因是cost不准确导致的，实际上这个文件对应的drop 回收站的记录只有1条！！！ 实际环境中并不一定就好使！
第二个delete sql explan 看看
EXPLAIN PLAN FOR delete from RecycleBin$
where
bo=90695;
SQL> SELECT * FROM TABLE(DBMS_XPLAN.DISPLAY);
------------------------------------------------------------------------------------
| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time |
------------------------------------------------------------------------------------
| 0 | DELETE STATEMENT | | 1 | 52 | 1 (0)| 00:00:01 |
| 1 | DELETE | RECYCLEBIN$ | | | | |
|* 2 | INDEX RANGE SCAN| RECYCLEBIN$_BO | 1 | 52 | 1 (0)| 00:00:01 |
------------------------------------------------------------------------------------
Predicate Information (identified by operation id):
---------------------------------------------------
2 - access("BO"=90695)
尝试第三种方式shrink ?
SQL> select segment_name,segment_type,bytes/1024/1024 from user_segments where segment_name like 'RECYCLEBIN%';
SEGMENT_NAME SEGMENT_TYPE BYTES/1024/1024
------------------------- ------------------ ---------------
RECYCLEBIN$ TABLE 480
RECYCLEBIN$_OBJ INDEX 160
RECYCLEBIN$_TS INDEX 136
RECYCLEBIN$_OWNER INDEX 144
RECYCLEBIN$_BO INDEX .0625
SQL> alter table RECYCLEBIN$ ENABLE ROW MOVEMENT;
SQL> ALTER TABLE RECYCLEBIN$ SHRINK SPACE COMPACT CASCADE;
ORA-10635: Invalid segment or tablespace type
SQL> select tablespace_name from dba_segments where segment_name='RECYCLEBIN$';
TABLESPACE_NAME
------------------------------
SYSTEM
SQL> select tablespace_name,CONTENTS,EXTENT_MANAGEMENT,SEGMENT_SPACE_MANAGEMENT from dba_tablespaces;
TABLESPACE_NAME CONTENTS EXTENT_MAN SEGMEN
------------------------------ ---------- ---------- ------
SYSTEM PERMANENT LOCAL MANUAL
SYSAUX PERMANENT LOCAL AUTO
UNDOTBS1 UNDO LOCAL MANUAL
TEMP TEMPORARY LOCAL MANUAL
USERS PERMANENT LOCAL AUTO
EXAMPLE PERMANENT LOCAL AUTO
shrink 不允许！ 那么只能move了！
SQL> alter table RECYCLEBIN$ disable ROW MOVEMENT;
SQL> alter table RECYCLEBIN$ move;
SQL> select segment_name,segment_type,bytes/1024/1024 from user_segments where segment_name like 'RECYCLEBIN%';
SEGMENT_NAME SEGMENT_TYPE BYTES/1024/1024
------------------------- ------------------ ---------------
RECYCLEBIN$_OBJ INDEX 160
RECYCLEBIN$_TS INDEX 136
RECYCLEBIN$_OWNER INDEX 144
RECYCLEBIN$_BO INDEX .0625
RECYCLEBIN$ TABLE .0625
可以发现索引并未变小！ 索引需要重建！
SQL> alter index RECYCLEBIN$_OBJ rebuild;
SQL> alter index RECYCLEBIN$_TS rebuild;
SQL> alter index RECYCLEBIN$_OWNER rebuild;
SQL> alter index RECYCLEBIN$_BO rebuild;
SQL> select segment_name,segment_type,bytes/1024/1024 from user_segments where segment_name like 'RECYCLEBIN%';
SEGMENT_NAME SEGMENT_TYPE BYTES/1024/1024
------------------------- ------------------ ---------------
RECYCLEBIN$_TS INDEX .0625
RECYCLEBIN$_OWNER INDEX .0625
RECYCLEBIN$_BO INDEX .0625
RECYCLEBIN$_OBJ INDEX .0625
RECYCLEBIN$ TABLE .0625
SQL> select status,count(*) from dba_indexes group by status;
STATUS COUNT(*)
-------- ----------
VALID 5000 N/A 100
再次模拟一次空间回收，本次不观察执行计划，只观察功能是否无异常即可。
SQL> drop table scott.tmp2 purge;
SQL> SELECT SUM(BYTES)/1024/1024 FROM DBA_FREE_SPACE WHERE TABLESPACE_NAME='TESTTBS';
SUM(BYTES)/1024/1024
--------------------
4
SQL> create table scott.tmp1 tablespace testtbs as select * from dba_objects where rownum<=10000;
insert into scott.tmp1 select * from scott.tmp1;
insert into scott.tmp1 select * from scott.tmp1;
ORA-01653: unable to extend table SCOTT.TMP1 by 128 in tablespace TESTTBS
SQL> alter database datafile '/11.2.0.4/app/oracle/tt/testtbs.dbf' resize 10m;
SQL> insert into scott.tmp1 select * from scott.tmp1; 20000 rows created. SQL> commit;
不影响！ 回收清除策略，如果业务不使用的情况下可以move 基表，重建索引！重新收集统计信息。
业务高峰期可以创建索引，收集统计信息，走索引也很快。
select segment_name,segment_type,bytes/1024/1024 from user_segments where segment_name like 'RECYCLEBIN%'
SEGMENT_NAME SEGMENT_TYPE BYTES/1024/1024------------------------- ------------------ ---------------RECYCLEBIN$ TABLE 480RECYCLEBIN$_OBJ INDEX 160RECYCLEBIN$_TS INDEX 136RECYCLEBIN$_OWNER INDEX 144