| René Nyffenegger's collection of things on the web | |
|
René Nyffenegger on Oracle - Most wanted - Feedback
|
Dumping a block in Oracle | ||
|
It is possible to peek into the datafiles and see what data they contain. In order to demonstrate that, I create
a table (dump_table), fill some data in it and dump it into a file for further inspecting.
The statement to dump one or more blocks is
alter system dump datafile <number datafile> block min <first number> block max <last block>
Here's the table that's going to be filled:
create table dump_table ( a varchar2(4000), b varchar2(4000), c varchar2(4000), d varchar2(4000)); insert into dump_table values( 'aaaaaaaaaa','bbbbbbbbbb','cccccccccc','dddddddddd'); insert into dump_table values( 'AAAAAAAAAA','BBBBBBBBBB','CCCCCCCCCC','DDDDDDDDDD'); commit; Dumping the Segment's Header Block
Each table (actually each segment) has a segment header.
First, the file and the block number must be determined before the block can be dumped
SQL> select header_file, header_block from dba_segments where segment_name = 'DUMP_TABLE'; HEADER_FILE HEADER_BLOCK ----------- ------------ 3 108639 alter system dump datafile 3 block 108639; Location of dumpfile
This command dumps the block in a more or less readable form. Where is it dumped? The
directory in which it is dumped is found like this:
select value from v$parameter where name = 'user_dump_dest';
The name of the dumped file is then determined like so:
select spid from v$session s, v$process p where p.addr = s.paddr and s.audsid = sys_context('userenv','sessionid')
This returns a number. Append this number to ORA and append a suffix .trc
See find the trace file for a complete script.
Here's the interesting part of the dump file:
Start dump data blocks tsn: 3 file#: 3 minblk 49 maxblk 49
buffer tsn: 3 rdba: 0x00c00031 (3/49)
scn: 0x0000.0005c24e seq: 0x01 flg: 0x04 tail: 0xc24e1001
frmt: 0x02 chkval: 0x49ed type: 0x10=DATA SEGMENT HEADER - UNLIMITED
Extent Control Header
-----------------------------------------------------------------
Extent Header:: spare1: 0 spare2: 0 #extents: 1 #blocks: 15
last map 0x00000000 #maps: 0 offset: 2080
Highwater:: 0x00c00033 ext#: 0 blk#: 1 ext size: 15
#blocks in seg. hdr's freelists: 1
#blocks below: 1
mapblk 0x00000000 offset: 0
Unlocked
Map Header:: next 0x00000000 #extents: 1 obj#: 5927 flag: 0x40000000
Extent Map
-----------------------------------------------------------------
0x00c00032 length: 15
nfl = 1, nfb = 1 typ = 1 nxf = 0 ccnt = 1
SEG LST:: flg: USED lhd: 0x00c00032 ltl: 0x00c00032
End dump data blocks tsn: 3 file#: 3 minblk 49 maxblk 49
The number after obj# (that is 5927) is found in sys.obj$:
select name from obj$ where obj# = 5927;
returns DUMP_TABLE.
Dumping a block from a rowid
The following procedure can be used to dump a database block that contains
a row whose rowid is known.
The user who compiles this procedure must have the rights to select
v$parameter,
v$process,
v$session, otherwise he'll get
an ora-00942 table or view does not exist.
create or replace procedure dump_block_from_rowid(p_rowid rowid) is dump_file varchar2(4000); begin execute immediate ' alter system dump datafile ' || dbms_rowid.rowid_relative_fno(p_rowid) || ' block ' || dbms_rowid.rowid_block_number(p_rowid); select u_dump.value || '/' || instance.value || '_ora_' || sys.v_$process.spid || '.trc' into dump_file from sys.v_$parameter u_dump cross join sys.v_$parameter instance cross join sys.v_$process join sys.v_$session on sys.v_$process.addr = sys.v_$session.paddr where u_dump.name = 'user_dump_dest' and instance.name = 'instance_name' and sys.v_$session.audsid=sys_context('userenv','sessionid'); dbms_output.put_line(' dumped block to: '); dbms_output.put_line(' ' || dump_file); end; / Other things
See also Dumping an index tree
|