[20181122]模拟ORA-08103错误.txt
[20181122]模拟ora-08103错误.txt
$ oerr ora 8103
08103, 00000, "object no longer exists"
// *cause: the object has been deleted by another user since the operation
// began, or a prior incomplete recovery restored the database to
// a point in time during the deletion of the object.
// *action: delete the object if this is the result of an incomplete
// recovery.
在 master note for handling oracle database corruption issues (文档 id 1088018.1)中对ora-8103错误的描述如下：
the object has been deleted by another user since the operation began.
if the error is reproducible, following may be the reasons:-
a.) the header block has an invalid block type.
b.) the data_object_id (seg/obj) stored in the block is different than the data_object_id stored in the segment header.
see dba_objects.data_object_id and compare it to the decimal value stored in the block (field seg/obj).
--//以前做过的测试.链接:http://blog.itpub.net/267265/viewspace-2131848/
--//听别人提起高水位下的块出现了未格式化的块.自己模拟测试看看.
1.环境:
scott@book> @ ver1
port_string version banner
------------------------------ -------------- --------------------------------------------------------------------------------
x86_64/linux 2.4.xx 11.2.0.4.0 oracle database 11g enterprise edition release 11.2.0.4.0 - 64bit production
--//建立表空间:
create tablespace tea datafile
'/mnt/ramdisk/book/tea01.dbf' size 40m autoextend on next 1m maxsize unlimited
logging
online
extent management local autoallocate
blocksize 8k
segment space management manual
flashback on;
--//顺便做1个备份.
rman> backup as copy datafile 6 format '/home/oracle/backup/%b' ;
starting backup at 2018-11-22 16:22:12
using target database control file instead of recovery catalog
allocated channel: ora_disk_1
channel ora_disk_1: sid=67 device type=disk
allocated channel: ora_disk_2
channel ora_disk_2: sid=94 device type=disk
allocated channel: ora_disk_3
channel ora_disk_3: sid=106 device type=disk
channel ora_disk_1: starting datafile copy
input datafile file number=00006 name=/mnt/ramdisk/book/tea01.dbf
output file name=/home/oracle/backup/tea01.dbf tag=tag20181122t162213 recid=13 stamp=992881334
channel ora_disk_1: datafile copy complete, elapsed time: 00:00:01
finished backup at 2018-11-22 16:22:15
starting control file and spfile autobackup at 2018-11-22 16:22:15
piece handle=/u01/app/oracle/fast_recovery_area/book/autobackup/2018_11_22/o1_mf_s_992881335_fzdssq4g_.bkp comment=none
finished control file and spfile autobackup at 2018-11-22 16:22:16
2.建立测试环境:
scott@book> create table t tablespace tea as select * from all_objects where rownum<=1e4;
table created.
scott@book> select object_id,data_object_id from dba_objects where object_name = 't' and owner = user;
object_id data_object_id
---------- --------------
90463 90463
scott@book> select segment_type, header_file, header_block from dba_segments where owner = user and segment_name = 't';
segment_type header_file header_block
------------------ ----------- ------------
table 6 128
scott@book> select file_id, block_id, blocks from dba_extents where owner = user and segment_name = 't';
file_id block_id blocks
---------- ---------- ----------
6 128 8
6 136 8
6 144 8
6 152 8
6 160 8
6 168 8
6 176 8
6 184 8
6 192 8
6 200 8
6 208 8
6 216 8
6 224 8
6 232 8
6 240 8
6 248 8
6 256 128
17 rows selected.
2.破坏数据块看看:
scott@book> alter system flush buffer_cache;
system altered.
scott@book> select max(rowid) from t ;
max(rowid)
------------------
aaawffaagaaaaebabj
scott@book> @ rowid aaawffaagaaaaebabj
object file block row rowid_dba dba text
---------- ---------- ---------- ---------- -------------------- -------------------- ----------------------------------------
90463 6 257 73 0x1800101 6,257 alter system dump datafile 6 block 257 ;
--//dba=6,257块上有数据.
scott@book> insert into t select * from all_objects where rownum<=1;
1 row created.
scott@book> commit ;
commit complete.
scott@book> alter system flush buffer_cache;
system altered.
scott@book> @ rowid aaawffaagaaaaecaaa
object file block row rowid_dba dba text
---------- ---------- ---------- ---------- -------------------- -------------------- ----------------------------------------
90463 6 258 0 0x1800102 6,258 alter system dump datafile 6 block 258 ;
--//dba=6,258块上有数据.通过bbed观察dba=6,259
bbed> map dba 6,259
file: /mnt/ramdisk/book/tea01.dbf (6)
block: 259 dba:0x01800103
------------------------------------------------------------
ktb data block (table/cluster)
struct kcbh, 20 bytes @0
struct ktbbh, 72 bytes @20
struct kdbh, 14 bytes @92
struct kdbt[0], 0 bytes @106
sb2 kdbr[0] @106
ub1 freespace[8082] @106
ub1 rowdata[0] @8188
ub4 tailchk @8188
--//可以发现dba=6,259已经格式化.但是没有数据.
scott@book> select count(*) from t;
count(*)
----------
10001
--//ok一切正常.现在破坏dba=6,259
bbed> set offset 0
offset 0
--//注意一定要设置offset 0,不然可能copy仅仅剩下的部分.
bbed> copy filename '/home/oracle/backup/tea01.dbf' block 259 to filename '/mnt/ramdisk/book/tea01.dbf' block 259
warning: contents of previous bifile will be lost. proceed? (y/n) y
file: /mnt/ramdisk/book/tea01.dbf (6)
block: 259 offsets: 0 to 63 dba:0x01800103
------------------------------------------------------------------------------------------------------------------------------------------------
00a20000 03018001 00000000 00000105 83a70000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000
<64 bytes per line>
--//现在已经未格式化块.
4.继续测试:
scott@book> alter system flush buffer_cache;
system altered.
scott@book> select count(*) from t;
select count(*) from t
*
error at line 1:
ora-08103: object no longer exists
--//再线ora-08103.
--//而执行如下不会报错:
scott@book> select count(*) from t where rownum<=10001;
count(*)
----------
10001
scott@book> select count(*) from t where rownum<=10002;
select count(*) from t where rownum<=10002
*
error at line 1:
ora-08103: object no longer exists
--//检查跟踪文件:
kcbzibmlt: dump suspect buffer, err=8103
buffer tsn: 7 rdba: 0x01800103 (6/259)
scn: 0x0000.00000000 seq: 0x01 flg: 0x05 tail: 0x00000001
frmt: 0x02 chkval: 0xa783 type: 0x00=unknown
hex dump of corrupt header 4 = corrupt
dump of memory from 0x000000006a4ea000 to 0x000000006a4ea014
06a4ea000 0000a200 01800103 00000000 05010000 [................]
06a4ea010 0000a783 [....]
hex dump of block: st=4, typ_found=0
dump of memory from 0x000000006a4ea000 to 0x000000006a4ec000
06a4ea000 0000a200 01800103 00000000 05010000 [................]
06a4ea010 0000a783 00000000 00000000 00000000 [................]
06a4ea020 00000000 00000000 00000000 00000000 [................]
repeat 508 times
06a4ebff0 00000000 00000000 00000000 00000001 [................]
dump of buffer cache at level 8 for tsn=7 rdba=25166080
--//可以发现dba =6,259报错.
5.如何跳过呢?
--//我测试仅仅设置坏块.如果真有数据可以通过一些特殊的方式读出来,这个给看运气.通过10231事件跳过坏块
$ oerr ora 10231
10231, 00000, "skip corrupted blocks on _table_scans_"
// *cause:
// *action: such blocks are skipped in table scans, and listed in trace files
--//alter session set db_file_multiblock_read_count=1 ;
alter session set events '10231 trace name context forever, level 10';
scott@book> alter session set events '10231 trace name context forever, level 10';
session altered.
scott@book> select count(*) from t ;
select count(*) from t
*
error at line 1:
ora-08103: object no longer exists
--//不行.在bbed设置坏块看看.
bbed> set dba 6,259
dba 0x01800103 (25166083 6,259)
bbed> corrupt
block marked media corrupt.
bbed> sum apply ;
check value for file 6, block 259:
current = 0xa683, required = 0xa683
scott@book> select count(*) from t ;
count(*)
----------
10001
--//ok.跳过坏块.