谈谈Oracle dba_free_space

顾名思义，dba_free_space指的是Oracle还有多少表空间剩余空间，其视图结构也相当简单：
SQL> desc dba_free_space
Name                                      Null?    Type
----------------------------------------- -------- ----------------------------
TABLESPACE_NAME                                    VARCHAR2(30)
FILE_ID                                            NUMBER
BLOCK_ID                                           NUMBER
BYTES                                              NUMBER
BLOCKS                                             NUMBER
RELATIVE_FNO                                       NUMBER
但是我们查询dba_free_space时，即表空间剩余空间常常是离碎的，比如
SQL> select * from dba_free_space where file_id=7;

TABLESPACE_NAME                   FILE_ID   BLOCK_ID      BYTES     BLOCKS RELATIVE_FNO
------------------------------ ---------- ---------- ---------- ---------- ------------
ZHOUL                                   7      27145     983040        120            7
ZHOUL                                   7      27905      65536          8            7
ZHOUL                                   7      28937    7274496        888            7
ZHOUL                                   7      36617     851968        104            7
ZHOUL                                   7      60129     327680         40            7
ZHOUL                                   7      63497     720896         88            7

6 rows selected.
这是为什么呢？继续查看视图dba_free_space的创建语句：
create or replace view dba_free_space
(tablespace_name, file_id, block_id, bytes, blocks, relative_fno)
as
select ts.name, fi.file#, f.block#,
       f.length * ts.blocksize, f.length, f.file#
from sys.ts$ ts, sys.fet$ f, sys.file$ fi
where ts.ts# = f.ts#
  and f.ts# = fi.ts#
  and f.file# = fi.relfile#
  and ts.bitmapped = 0
union all
select /*+ ordered use_nl(f) use_nl(fi) */
       ts.name, fi.file#, f.ktfbfebno,
       f.ktfbfeblks * ts.blocksize, f.ktfbfeblks, f.ktfbfefno
from sys.ts$ ts, sys.x$ktfbfe f, sys.file$ fi
where ts.ts# = f.ktfbfetsn
  and f.ktfbfetsn = fi.ts#
  and f.ktfbfefno = fi.relfile#
  and ts.bitmapped <> 0 and ts.online$ in (1,4) and ts.contents$ = 0
union all
select /*+ ordered use_nl(u) use_nl(fi) */
       ts.name, fi.file#, u.ktfbuebno,
       u.ktfbueblks * ts.blocksize, u.ktfbueblks, u.ktfbuefno
from sys.recyclebin$ rb, sys.ts$ ts, sys.x$ktfbue u, sys.file$ fi
where ts.ts# = rb.ts#
  and rb.ts# = fi.ts#
  and rb.file# = fi.relfile#
  and u.ktfbuesegtsn = rb.ts#
  and u.ktfbuesegfno = rb.file#
  and u.ktfbuesegbno = rb.block#
  and ts.bitmapped <> 0 and ts.online$ in (1,4) and ts.contents$ = 0
union all
select ts.name, fi.file#, u.block#,
       u.length * ts.blocksize, u.length, u.file#
from sys.ts$ ts, sys.uet$ u, sys.file$ fi, sys.recyclebin$ rb
where ts.ts# = u.ts#
  and u.ts# = fi.ts#
  and u.segfile# = fi.relfile#
  and u.ts# = rb.ts#
  and u.segfile# = rb.file#
  and u.segblock# = rb.block#
  and ts.bitmapped = 0;
可以看到dba_free_space视图有三部分组成：fet$，x$ktfbfe，x$ktfbue，recyclebin$。其中fet$表格主要用于表空间extent管理是数据字典管理，x$ktfbue由前面的实验得知主要用于对位图块的扫描，recyclebin$主要用于管理回收站对象。
那x$ktfbfe主要用于做什么呢？
View:   X$KTFBUE
         [k]ernel [t]ablespace [f]ile [b]itmapped
           [u]sed [e]xtents


Column          Type               Description
--------        ----               --------
ADDR            RAW(4|8)           address of this row/entry in the array or SGA
INDX            NUMBER             index number of this row in the fixed table array
INST_ID         NUMBER             oracle instance number
KTFBUESEGTSN    NUMBER             tablespace number of segment
KTFBUESEGFNO    NUMBER             segment relative file number
KTFBUESEGBNO    NUMBER             segment block number
KTFBUEEXTNO     NUMBER             extent number
KTFBUEFNO       NUMBER             extent file number
KTFBUEBNO       NUMBER             extent block number
KTFBUEBLKS      NUMBER             extent length
打开10046事件跟踪x$ktfbfe

SQL> ALTER SESSION SET EVENTS '10046 trace name context forever, level 12';

Session altered.

SQL> select * from x$ktfbfe;

ADDR           INDX    INST_ID  KTFBFETSN  KTFBFEFNO  KTFBFEBNO KTFBFEBLKS
-------- ---------- ---------- ---------- ---------- ---------- ----------
B7F57A70          0          1          0          1      69769        632
B7F57A70          1          1          1          2        233      18328
B7F57A70          2          1          2          3      36953         80
B7F57A70          3          1          2          3      37041         40
B7F57A70          4          1          2          3      37121          8
。。。
SQL> ALTER SESSION SET EVENTS '10046 trace name context off';

Session altered.
打开跟踪文件，首先Oracle对x$ktfbfe进行解析
PARSING IN CURSOR #1 len=22 dep=0 uid=0 oct=3 lid=0 tim=1273325024428885 hv=502180737 ad='2674fde8'
select * from x$ktfbfe
END OF STMT
PARSE #1:c=0,e=130,p=0,cr=0,cu=0,mis=0,r=0,dep=0,og=1,tim=1273325024428880
BINDS #1:
EXEC #1:c=0,e=78,p=0,cr=0,cu=0,mis=0,r=0,dep=0,og=1,tim=1273325024429032
WAIT #1: nam='SQL*Net message to client' ela= 1 driver id=1650815232 #bytes=1 p3=0 obj#=-1 tim=1273325024429077
其次Oracle进一步解析ts$表，获取满足条件的ts#和flags
PARSING IN CURSOR #2 len=100 dep=1 uid=0 oct=3 lid=0 tim=1273325024429259 hv=3768030067 ad='25b84394'
select ts#, flags from ts$ where bitmapped <> 0 and contents$ = 0   and (online$ = 1 or online$ = 4)
END OF STMT
PARSE #2:c=0,e=91,p=0,cr=0,cu=0,mis=0,r=0,dep=1,og=4,tim=1273325024429255
BINDS #2:
EXEC #2:c=1000,e=50,p=0,cr=0,cu=0,mis=0,r=0,dep=1,og=4,tim=1273325024429371
WAIT #2: nam='db file sequential read' ela= 39 file#=1 block#=57 blocks=1 obj#=16 tim=1273325024442859
WAIT #2: nam='db file scattered read' ela= 130 file#=1 block#=58 blocks=7 obj#=16 tim=1273325024443240
FETCH #2:c=1000,e=13908,p=8,cr=4,cu=0,mis=0,r=1,dep=1,og=4,tim=1273325024443299
最后Oracle解析file$，传入绑定变量0,1,2...8
PARSING IN CURSOR #3 len=36 dep=1 uid=0 oct=3 lid=0 tim=1273325024443480 hv=1570213724 ad='27af1440'
select file# from file$ where ts#=:1
END OF STMT
PARSE #3:c=0,e=80,p=0,cr=0,cu=0,mis=0,r=0,dep=1,og=4,tim=1273325024443476
BINDS #3:
kkscoacd
Bind#0
  oacdty=02 mxl=22(22) mxlc=00 mal=00 scl=00 pre=00
  oacflg=08 fl2=0001 frm=00 csi=00 siz=24 off=0
  kxsbbbfp=b7f65180  bln=22  avl=01  flg=05
  value=0
EXEC #3:c=0,e=119,p=0,cr=0,cu=0,mis=0,r=0,dep=1,og=4,tim=1273325024443693
WAIT #3: nam='db file sequential read' ela= 14835 file#=1 block#=113 blocks=1 obj#=16 tim=1273325024458576
WAIT #3: nam='db file sequential read' ela= 236 file#=1 block#=114 blocks=1 obj#=16 tim=1273325024458882
FETCH #3:c=999,e=15195,p=2,cr=3,cu=0,mis=0,r=1,dep=1,og=4,tim=1273325024458914
FETCH #3:c=0,e=8,p=0,cr=1,cu=0,mis=0,r=0,dep=1,og=4,tim=1273325024458949
STAT #3 id=1 cnt=1 pid=0 pos=1 obj=17 op='TABLE ACCESS FULL FILE$ (cr=4 pr=2 pw=0 time=15194 us)'
WAIT #1: nam='db file sequential read' ela= 16 file#=1 block#=2 blocks=1 obj#=-1 tim=1273325024459119
WAIT #1: nam='db file sequential read' ela= 15 file#=1 block#=3 blocks=1 obj#=-1 tim=1273325024459190
FETCH #1:c=2999,e=30138,p=12,cr=8,cu=2,mis=0,r=1,dep=0,og=1,tim=1273325024459250
WAIT #1: nam='SQL*Net message from client' ela= 244 driver id=1650815232 #bytes=1 p3=0 obj#=-1 tim=1273325024459537
FETCH #2:c=0,e=9,p=0,cr=1,cu=0,mis=0,r=1,dep=1,og=4,tim=1273325024459584
。。。
PARSING IN CURSOR #3 len=36 dep=1 uid=0 oct=3 lid=0 tim=1273325024482416 hv=1570213724 ad='27af1440'
select file# from file$ where ts#=:1
END OF STMT
PARSE #3:c=0,e=13,p=0,cr=0,cu=0,mis=0,r=0,dep=1,og=4,tim=1273325024482412
BINDS #3:
kkscoacd
Bind#0
  oacdty=02 mxl=22(22) mxlc=00 mal=00 scl=00 pre=00
  oacflg=08 fl2=0001 frm=00 csi=00 siz=24 off=0
  kxsbbbfp=b7f65134  bln=22  avl=02  flg=05
  value=7
EXEC #3:c=0,e=92,p=0,cr=0,cu=0,mis=0,r=0,dep=1,og=4,tim=1273325024482587
FETCH #3:c=0,e=22,p=0,cr=3,cu=0,mis=0,r=1,dep=1,og=4,tim=1273325024482634
FETCH #3:c=0,e=5,p=0,cr=1,cu=0,mis=0,r=0,dep=1,og=4,tim=1273325024482666
STAT #3 id=1 cnt=1 pid=0 pos=1 obj=17 op='TABLE ACCESS FULL FILE$ (cr=4 pr=0 pw=0 time=19 us)'
WAIT #1: nam='db file sequential read' ela= 17 file#=6 block#=2 blocks=1 obj#=-1 tim=1273325024482802
WAIT #1: nam='db file sequential read' ela= 15 file#=6 block#=3 blocks=1 obj#=-1 tim=1273325024482875
FETCH #2:c=0,e=7,p=0,cr=1,cu=0,mis=0,r=1,dep=1,og=4,tim=1273325024482917

通过查看跟踪文件，我们可以看到Oracle对x$ktfbfe表格的查询，最终会转换成对ts$的查询，通过条件过滤定位file$，然后从文件的2号block和3号block去取得数据。
我们知道每个数据文件的2号至-8号block是关于extent map的block。
从block type为1d可以知道这个block类型为KTFB Bitmapped File Space Header
BBED> dump block 2 offset 0 count 32
File: /oradata/mcstar/zhoul01.dbf (0)
Block: 2                Offsets:    0 to   31           Dba:0x00000000
------------------------------------------------------------------------
1da20000 0200c001 1fb3840e 000a0304 15fd0000 07000000 08000000 60f80000

<32 bytes per line>

从block type为12可以知道这个block类型为KTFB Bitmapped File Space Bitmap
BBED> dump block 3 offset 0 count 32
File: /oradata/mcstar/zhoul01.dbf (0)
Block: 3                Offsets:    0 to   31           Dba:0x00000000
------------------------------------------------------------------------
1ea20000 0300c001 1fb3840e 000a0104 35cd0000 07000000 09000000 00000000

<32 bytes per line>

从以上分析中我们推断出，Oracle查看x$ktfbfe，其实就是对Oracle 数据文件的block 2至block 8扫描（本例block 3-8为空，则跳过不扫描）。
从dba_free_space视图创建脚本中，我们还看到了表格recyclebin$内容的选取
继续测试：
在数据库中删除一张表格RBOTEST，其数据量有52567
SQL> select count(*) from RBOTEST;

  COUNT(*)
----------
     52567

SQL> drop table RBOTEST;

Table dropped.
刷内存，保证脏块刷出至数据文件
SQL>  alter system flush buffer_cache;

System altered.
在recyclebin中我们看到了删除表格
SQL> show recyclebin
ORIGINAL NAME    RECYCLEBIN NAME                OBJECT TYPE  DROP TIME
---------------- ------------------------------ ------------ -------------------
RBOTEST          BIN$oeDriA+aATTgQBCsowQS+Q==$0 TABLE        2011-04-27:14:55:03
在基表中也存在
SQL> select OBJ#,OWNER#,ORIGINAL_NAME,FILE#,BLOCK# ,FLAGS,SPACE from recyclebin$;

      OBJ#     OWNER# ORIGINAL_NAME                         FILE#     BLOCK#      FLAGS      SPACE
---------- ---------- -------------------------------- ---------- ---------- ---------- ----------
    246366         60 RBOTEST_OBJ                               7      29579         18        128
    246367         60 RBOTEST_OWNER#                            7      29707         18        112
    246365         60 RBOTEST                                   7      27147         30        768
但是在x$ktfbfe显示依然是删除前的状态
SQL> select * from x$ktfbfe
  2  where ktfbfefno=7;

ADDR           INDX    INST_ID  KTFBFETSN  KTFBFEFNO  KTFBFEBNO KTFBFEBLKS
-------- ---------- ---------- ---------- ---------- ---------- ----------
B7F57A70         50          1          8          7      29817          8
B7F57A70         51          1          8          7      36617        104
B7F57A70         52          1          8          7      60129         40
B7F57A70         53          1          8          7      63497         88

通过bbed查看block状态，发现checkval值未变，这说明Oracle在Oracle 10g中drop 表格时extent map并未发生变化
BBED> dump block 2 offset 0 count 32
File: /oradata/mcstar/zhoul01.dbf (0)
Block: 2                Offsets:    0 to   31           Dba:0x00000000
------------------------------------------------------------------------
1da20000 0200c001 1fb3840e 000a0304 15fd0000 07000000 08000000 60f80000

<32 bytes per line>


BBED> dump block 3 offset 0 count 32
File: /oradata/mcstar/zhoul01.dbf (0)
Block: 3                Offsets:    0 to   31           Dba:0x00000000
------------------------------------------------------------------------
1ea20000 0300c001 1fb3840e 000a0104 35cd0000 07000000 09000000 00000000

<32 bytes per line>
清空回收站
SQL> purge recyclebin;

Recyclebin purged.

SQL> select OBJ#,OWNER#,ORIGINAL_NAME,FILE#,BLOCK# ,FLAGS,SPACE from recyclebin$;

no rows selected

继续查看x$ktfbfe和物理上block状态，发现未变，继续刷内存。
SQL> alter system flush buffer_cache;

System altered.

SQL> select * from x$ktfbfe
  2  where ktfbfefno=7;

ADDR           INDX    INST_ID  KTFBFETSN  KTFBFEFNO  KTFBFEBNO KTFBFEBLKS
-------- ---------- ---------- ---------- ---------- ---------- ----------
B7F26A58         47          1          8          7      27145        120
B7F26A58         48          1          8          7      27905          8
B7F26A58         49          1          8          7      28937        888
B7F26A58         50          1          8          7      36617        104
B7F26A58         51          1          8          7      60129         40
B7F26A58         52          1          8          7      63497         88

6 rows selected.

BBED> dump block 2 offset 0 count 32
File: /oradata/mcstar/zhoul01.dbf (0)
Block: 2                Offsets:    0 to   31           Dba:0x00000000
------------------------------------------------------------------------
1da20000 0200c001 5879930e 000a0104 955b0000 07000000 08000000 60f80000

<32 bytes per line>

BBED> dump block 3 offset 0 count 32
File: /oradata/mcstar/zhoul01.dbf (0)
Block: 3                Offsets:    0 to   31           Dba:0x00000000
------------------------------------------------------------------------
1ea20000 0300c001 5879930e 000a0104 6e0e0000 07000000 09000000 00000000

<32 bytes per line>

可以看到x$ktfbfe和block均发生了变化。
通过对dba_free_space的研究我们可以得出以下结论：
1、对x$ktfbfe的扫描，其实是物理上对数据文件2-8号block的扫描
2、在Oracle 10g，在不带参数purge，drop表格时，并不会对数据文件头extent map更新，通过这种方式减少了Oracle对extent map争用的可能性，这也是dba_free_space视图创建脚本中需要对基表recyclebin$选择的原因之一。曾经碰到过一案例，回收站对象太多导致执行dba_free_space时间很长。
3、在对回收站清空后，会更新数据文件头extent map，但命令purge recyclebin并不会引起对象基表的checkpoint。