[20150811]模擬壞塊處理.txt

[20150811]模擬壞塊處理.txt

--如果存在備份,修復壞塊還是相對簡單的.在11g下:

select * from V$DATABASE_BLOCK_CORRUPTION;

--在rman下執行:
blockrecover corruption list;

--如果資料塊沒有使用,沒有分配data_object_id而出現壞塊,如何恢復呢?一般採用的方法建立新物件的方法,格式化這個資料塊.
--具體測試如下:

1.建立測試環境:

SCOTT@test> @ver1
PORT_STRING                    VERSION        BANNER
------------------------------ -------------- ----------------------------------------------------------------
x86_64/Linux 2.4.xx            10.2.0.4.0     Oracle Database 10g Enterprise Edition Release 10.2.0.4.0 - 64bi

CREATE TABLESPACE MSSM DATAFILE
  '/mnt/ramdisk/test/mssm01.dbf' SIZE 64M AUTOEXTEND OFF
LOGGING
ONLINE
EXTENT MANAGEMENT LOCAL AUTOALLOCATE
BLOCKSIZE 8K
SEGMENT SPACE MANAGEMENT MANUAL
FLASHBACK ON;

SCOTT@test> create table t tablespace mssm as select rownum id ,cast('testtesttesttest' as varchar2(20)) name from xmltable('1 to 400000');
Table created.

--這樣建立檔案大小12M。

--建立備份:
backup database format '/home/oracle/backup/full_%u';
backup datafile 6 format '/home/oracle/backup/DATAFILE6_%u' ;
backup as copy datafile 6  format '/home/oracle/backup/mssm01.dbf_copy' ;

$  cd /home/oracle/backup
$  mv mssm01.dbf_copy mssm01.dbf_copy_org
--delete force copy of datafile 6;

--修改block＝3000的資訊：
--3000*8192/1024/1024=23.4375，這樣沒有資訊寫在該塊。

2.關閉資料庫,破壞資料塊:

SCOTT@test> @bbvi 6 3000
BVI_COMMAND
------------------------------------------------------
bvi -b 24576000 -s 8192 /mnt/ramdisk/test/mssm01.dbf

SCOTT@test> @convrdba.sql 6 3000
RDBA16                 RDBA
-------------- ------------
       1800bb8     25168824

$ bvi -b 24576000 -s 8192 /mnt/ramdisk/test/mssm01.dbf
01770000  00 A2 00 00 B8 0B 00 00 00 00 00 00 00 00 01 05 B8 AC 0
--將這些資訊清零。
--我使用bvi,使用dd也可以.注意要加conv=notrunc引數.方向不要搞錯。
--dd if=/dev/zero of=/mnt/ramdisk/test/mssm01.dbf bs=8192 seek=3000 conv=notrunc count=1

$  dbv file=/mnt/ramdisk/test/mssm01.dbf
DBVERIFY: Release 10.2.0.4.0 - Production on Tue Aug 11 08:25:08 2015
Copyright (c) 1982, 2007, Oracle.  All rights reserved.
DBVERIFY - Verification starting : FILE = /mnt/ramdisk/test/mssm01.dbf
Page 3000 is influx - most likely media corrupt
Corrupt block relative dba: 0x01800bb8 (file 6, block 3000)
Fractured block found during dbv:
Data in bad block:
type: 0 format: 0 rdba: 0x00000000
last change scn: 0x0000.00000000 seq: 0x0 flg: 0x00
spare1: 0x0 spare2: 0x0 spare3: 0x0
consistency value in tail: 0x00000001
check value in block header: 0x0
block checksum disabled

DBVERIFY - Verification complete
Total Pages Examined         : 8192
Total Pages Processed (Data) : 1492
Total Pages Failing   (Data) : 0
Total Pages Processed (Index): 0
Total Pages Failing   (Index): 0
Total Pages Processed (Other): 9
Total Pages Processed (Seg)  : 0
Total Pages Failing   (Seg)  : 0
Total Pages Empty            : 6690
Total Pages Marked Corrupt   : 1
Total Pages Influx           : 1
Highest block SCN            : 4147024117 (2.4147024117)
--可以發現Corrupt block relative dba: 0x01800bb8 (file 6, block 3000)資訊，表示存在壞塊。

SCOTT@test> select * from V$DATABASE_BLOCK_CORRUPTION;
no rows selected
--可以發現檢視V$DATABASE_BLOCK_CORRUPTION沒有記錄。

RMAN> backup validate datafile 6;
Starting backup at 2015-08-11 08:26:44
using target database control file instead of recovery catalog
allocated channel: ORA_DISK_1
channel ORA_DISK_1: sid=143 devtype=DISK
channel ORA_DISK_1: starting full datafile backupset
channel ORA_DISK_1: specifying datafile(s) in backupset
input datafile fno=00006 name=/mnt/ramdisk/test/mssm01.dbf
channel ORA_DISK_1: backup set complete, elapsed time: 00:00:01
Finished backup at 2015-08-11 08:26:45

SCOTT@test> select * from V$DATABASE_BLOCK_CORRUPTION;
no rows selected
--10g下對這種情況沒有記錄。使用blockrecover corruption list;應該沒用。使用dbv檢查依舊。

RMAN> blockrecover corruption list;
Starting blockrecover at 2015-08-11 08:27:55
using channel ORA_DISK_1
starting media recovery
media recovery complete, elapsed time: 00:00:00
Finished blockrecover at 2015-08-11 08:27:55

--直接指定資料檔案以及對應塊。
RMAN> blockrecover datafile 6 block 3000 ;
Starting blockrecover at 2015-08-11 08:29:32
using channel ORA_DISK_1

channel ORA_DISK_1: restoring block(s)
channel ORA_DISK_1: specifying block(s) to restore from backup set
restoring blocks of datafile 00006
channel ORA_DISK_1: reading from backup piece /home/oracle/backup/DATAFILE6_10qeakdl
channel ORA_DISK_1: restored block(s) from backup piece 1
piece handle=/home/oracle/backup/DATAFILE6_10qeakdl tag=TAG20150811T081133
channel ORA_DISK_1: block restore complete, elapsed time: 00:00:01

starting media recovery
media recovery complete, elapsed time: 00:00:03

Finished blockrecover at 2015-08-11 08:29:36

$  dbv file=/mnt/ramdisk/test/mssm01.dbf
DBVERIFY: Release 10.2.0.4.0 - Production on Tue Aug 11 08:30:21 2015
Copyright (c) 1982, 2007, Oracle.  All rights reserved.
DBVERIFY - Verification starting : FILE = /mnt/ramdisk/test/mssm01.dbf
DBVERIFY - Verification complete
Total Pages Examined         : 8192
Total Pages Processed (Data) : 1492
Total Pages Failing   (Data) : 0
Total Pages Processed (Index): 0
Total Pages Failing   (Index): 0
Total Pages Processed (Other): 10
Total Pages Processed (Seg)  : 0
Total Pages Failing   (Seg)  : 0
Total Pages Empty            : 6690
Total Pages Marked Corrupt   : 0
Total Pages Influx           : 0
Highest block SCN            : 4147024117 (2.4147024117)

--可以發現在有備份的情況下恢復實際上還是很簡單的，因為V$DATABASE_BLOCK_CORRUPTION沒有記錄，使用blockrecover corruption list;不行。
--但是直接執行blockrecover datafile 6 block 3000 ;，還是能修復問題的。

3.重複採用別的方式：
--採用rman的方式破壞。這種方式的破壞是往塊裡面寫入1堆垃圾(透過bvi觀察)，注意意後面的引數clear：
RMAN> blockrecover datafile 6 block 3000 clear ;
Starting blockrecover at 2015-08-11 08:37:51
using target database control file instead of recovery catalog
allocated channel: ORA_DISK_1
channel ORA_DISK_1: sid=157 devtype=DISK
Finished blockrecover at 2015-08-11 08:37:51

$  dbv file=/mnt/ramdisk/test/mssm01.dbf
DBVERIFY: Release 10.2.0.4.0 - Production on Tue Aug 11 08:39:02 2015
Copyright (c) 1982, 2007, Oracle.  All rights reserved.
DBVERIFY - Verification starting : FILE = /mnt/ramdisk/test/mssm01.dbf
Page 3000 is marked corrupt
Corrupt block relative dba: 0x01800bb8 (file 6, block 3000)
Bad check value found during dbv:
Data in bad block:
type: 58 format: 2 rdba: 0x01800bb8
last change scn: 0x0002.f72e9086 seq: 0x1 flg: 0x04
spare1: 0x0 spare2: 0x0 spare3: 0x0
consistency value in tail: 0x90863a01
check value in block header: 0x612e
computed block checksum: 0xef7c

-這種情況下，如果做如下操作會報錯：
RMAN> backup as copy datafile 6  format '/home/oracle/backup/mssm01.dbf_copy' ;

Starting backup at 2015-08-11 08:40:27
using channel ORA_DISK_1
channel ORA_DISK_1: starting datafile copy
input datafile fno=00006 name=/mnt/ramdisk/test/mssm01.dbf
RMAN-00571: ===========================================================
RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
RMAN-00571: ===========================================================
RMAN-03009: failure of backup command on ORA_DISK_1 channel at 08/11/2015 08:40:28
ORA-19566: exceeded limit of 0 corrupt blocks for file /mnt/ramdisk/test/mssm01.dbf

SCOTT@test> select * from V$DATABASE_BLOCK_CORRUPTION;
no rows selected

--依舊沒有記錄。因為在備份資料檔案時使用copy方式要檢查每個塊，而backup datafile 6 僅僅備份有資訊的塊，這樣使用copy方式會
--報錯。

SCOTT@test> create table tx tablespace mssm as select * from t where 1=2;
Table created.

SCOTT@test> alter table tx allocate extent  (size 20M);
Table altered.

$  dbv file=/mnt/ramdisk/test/mssm01.dbf
DBVERIFY: Release 10.2.0.4.0 - Production on Tue Aug 11 08:48:25 2015
Copyright (c) 1982, 2007, Oracle.  All rights reserved.
DBVERIFY - Verification starting : FILE = /mnt/ramdisk/test/mssm01.dbf
Page 3000 is marked corrupt
Corrupt block relative dba: 0x01800bb8 (file 6, block 3000)
Bad check value found during dbv:
Data in bad block:
type: 58 format: 2 rdba: 0x01800bb8
last change scn: 0x0002.f72e9086 seq: 0x1 flg: 0x04
spare1: 0x0 spare2: 0x0 spare3: 0x0
consistency value in tail: 0x90863a01
check value in block header: 0x612e
computed block checksum: 0xef7c

RMAN> backup datafile 6 format '/home/oracle/backup/DATAFILE6_%u' ;
Starting backup at 2015-08-11 08:51:53
using channel ORA_DISK_1
channel ORA_DISK_1: starting full datafile backupset
channel ORA_DISK_1: specifying datafile(s) in backupset
input datafile fno=00006 name=/mnt/ramdisk/test/mssm01.dbf
channel ORA_DISK_1: starting piece 1 at 2015-08-11 08:51:53
RMAN-00571: ===========================================================
RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
RMAN-00571: ===========================================================
RMAN-03009: failure of backup command on ORA_DISK_1 channel at 08/11/2015 08:51:54
ORA-19566: exceeded limit of 0 corrupt blocks for file /mnt/ramdisk/test/mssm01.dbf
--這次報錯。因為該塊已經被tx佔用。

SCOTT@test> select * from V$DATABASE_BLOCK_CORRUPTION;
no rows selected
--可以發現這樣檢視V$DATABASE_BLOCK_CORRUPTION沒有記錄。

RMAN> backup validate datafile 6;
Starting backup at 2015-08-11 08:52:10
using channel ORA_DISK_1
channel ORA_DISK_1: starting full datafile backupset
channel ORA_DISK_1: specifying datafile(s) in backupset
input datafile fno=00006 name=/mnt/ramdisk/test/mssm01.dbf
channel ORA_DISK_1: backup set complete, elapsed time: 00:00:01
Finished backup at 2015-08-11 08:52:11

SCOTT@test> select * from V$DATABASE_BLOCK_CORRUPTION;
       FILE#       BLOCK#       BLOCKS CORRUPTION_CHANGE# CORRUPTIO
------------ ------------ ------------ ------------------ ---------
           6         3000            1                  0 CHECKSUM

--這次有記錄了。

4.選擇生成新資料覆蓋壞塊：
SCOTT@test> insert into tx values (null,null);
1 row created.

SCOTT@test> insert into tx values (null,null);
1 row created.

SCOTT@test> commit ;
Commit complete.

SCOTT@test> ALTER TABLE tx MINIMIZE RECORDS_PER_BLOCK;
Table altered.

--3000*8192/1024/1024=23.4375, 24M位置。前面已經佔用12M。
--(24-12)*1024*1024/8192=1536.
--這樣寫1536*2=3072條記錄基本就覆蓋有問題的資料塊。而且這樣寫日誌相對較小。

SCOTT@test> insert into tx select null,null from dual connect by level<=3100;
3100 rows created.

SCOTT@test> commit ;
Commit complete.

SCOTT@test> alter system checkpoint;
System altered.

$  dbv file=/mnt/ramdisk/test/mssm01.dbf
DBVERIFY: Release 10.2.0.4.0 - Production on Tue Aug 11 09:01:10 2015
Copyright (c) 1982, 2007, Oracle.  All rights reserved.
DBVERIFY - Verification starting : FILE = /mnt/ramdisk/test/mssm01.dbf
DBVERIFY - Verification complete
Total Pages Examined         : 8192
Total Pages Processed (Data) : 3045
Total Pages Failing   (Data) : 0
Total Pages Processed (Index): 0
Total Pages Failing   (Index): 0
Total Pages Processed (Other): 10
Total Pages Processed (Seg)  : 0
Total Pages Failing   (Seg)  : 0
Total Pages Empty            : 5137
Total Pages Marked Corrupt   : 0
Total Pages Influx           : 0
Highest block SCN            : 4147026312 (2.4147026312)

SCOTT@test> set null NULL
SCOTT@test> select rowid,tx.* from tx where DBMS_ROWID.ROWID_BLOCK_NUMBER (rowid)=3000;
ROWID                        ID NAME
------------------ ------------ --------------------
AAAQCjAAGAAAAu4AAA NULL         NULL
AAAQCjAAGAAAAu4AAB NULL         NULL

--可以發現已經修復。