Oracle数据库块的物理损坏与逻辑损坏

笔者认为不能简单的说物理硬盘的损坏导致块的物理损坏，Oracle bug导致块的逻辑损坏。这篇文章我们来详细讨论一下Oracle数据库的物理损坏以及逻辑损坏的概念。下面是一篇METALINK的文章：
H Rw!Gb0

Physical and Logical Block Corruptions. All you wanted to know about it. [ID 840978.1]
		修改时间 21-MAY-2012 类型 BULLETIN 状态 PUBLISHED

In this DocumentITPUB个人空间/ht6A7m;MU

Purpose

Scope

Details

Physical Block Corruptions

Fractured Block

Bad Checksum

Block Misplaced

Logical Block Corruptions

References

---

Applies to:

Oracle Server - Enterprise Edition - Version 9.2.0.1 to 11.2.0.3 [Release 9.2 to 11.2]
%hG5L;|-cnv0Information in this document applies to any platform.
N;`Hgvd,F y1H3j0***Checked for relevance on 21-May-2012*** ITPUB个人空间-[d0k)[[K)z#y
ITPUB个人空间3Pd T#|\*F)gTS&k

3Yv{#y#y0

Purpose

Oracle classifies the Data File Block corruptions as Physical and Logical. This is also referred as intra block corruptions. This document is intended to provide detailed information and errors example about it.

Scope

This document is intended for Database Administrators.

Details

Physical Block Corruptions

This kind of block corruptions are normally reported by Oracle with error ORA-1578 and the detailed corruption description is printed in the alert log.
~SP"t Km@P0
H+?-TiGfz2b0Corruption Examples are:

• Bad header - the beginning of the block (cache header) is corrupt with invalid values
• The block is Fractured/Incomplete - header and footer of the block do not match
• The block checksum is invalid
• The block is misplaced
• Zeroed out blocks / ORA-8103

Detailed Corruption Description:

Fractured Block

A Fractured block means that the block is incomplete. Information from the block header does not match the block tail.

Corrupt block relative dba: 0x0380e573 (file 14, block 58739)
-[)c3`*FVfQqqBY0 Fractured block found during buffer read ITPUB个人空间E T0Vu0S'XA7O4X2]
Data in bad block -
^Ln@1jkja'wp%D0 type: 6 format: 2 rdba: 0x0380e573 ITPUB个人空间7D#X2v(Eq9]3}%H
last change scn: 0x0288.8e5 a2f78 seq: 0x1 flg: 0x04 ITPUB个人空间-c!q7g1n9i5q1x
consistency value in tail: 0x00780601
g[7hD!i Br2r0check value in block header: 0x8739, computed block checksum: 0x2f00
/l&V#@9U~o-re0spare1: 0x0, spare2: 0x0, spare3: 0x0 ITPUB个人空间|}!MD*m$A"ew?!s
***
}'XA6?kR0Reread of rdba: 0x0380e573 (file 14, block 58739) found same corrupted data

ITPUB个人空间w5XK4f,J%`

Bad Checksum

Block Checksums are used to identify if the block was changed by something external to Oracle and after the block was last written by Oracle.
? kz }].C0Checksum is calculated by DBWR or direct loader before writing the block to disk and stored in the block header. Every time that the block is read and if db_block_checksum is different than false, Oracle calculates a checksum and compares it to the one stored in the block header. Reference Note 30706.1ITPUB个人空间VQ$tlO
ITPUB个人空间1U"q5p%e1m:Q6NO
Example of a corrupt block due to invalid checksum:

Corrupt block relative dba: 0x0380a58f (file 14, block 42383)
hW(x&tyY0 Bad check value found during buffer read
wN%}@#gDt2jy9[v9C0Data in bad block -
k YGF[(l(ZN.m!Q0type: 6 format: 2 rdba: 0x0380a58f ITPUB个人空间5x*xd#{ sz
last change scn: 0x0288.7784c5ee seq: 0x1 flg: 0x06
$r^\.sq6I7[\0consistency value in tail: 0xc5ee0601
!v!|%gu(Fr|;E0check value in block header: 0x68a7, computed block checksum: 0x2f00
Q{%J6P/d loRG0spare1: 0x0, spare2: 0x0, spare3: 0x0
T4U'XVgh.qX0***
mr9T Z$b2n}#h0Reread of rdba: 0x0380a58f (file 14, block 42383) found same corrupted data

]o+\w@B%M0A value different than zero (0x0) in "computed block checksum" means that the checksum
#uv-q5@&qf0@(^$|#m0differs and the result of this comparison is printed. ITPUB个人空间J4y"T3_vA&{b1]

6z;q%UD$Y*Cd o0

Block Misplaced

This is when Oracle detected that the content of the block being read belongs to a different block and the checksum is valid:

Corrupt block relative dba: 0x0d805a89 (file 54, block 23177)
O E8B3WA:r#Z{e0Bad header found during buffer read
&NG W&? n2ZB0Data in bad block -
f:JJY:h6a:j_0type: 6 format: 2 rdba: 0x0d805b08 ----&gt Block is different than expected 0x0d805a89 ITPUB个人空间Qo`h0^ J\,o8I/a
last change scn: 0x0692.86dc08e3 seq: 0x1 flg: 0x04 ITPUB个人空间(W-[\&N HD*w;a8zD s
consistency value in tail: 0x08e30601
M9{{@#V8I;r%O0check value in block header: 0x2a6e, computed block checksum: 0x0
f0MN&i6t0spare1: 0x0, spare2: 0x0, spare3: 0x0
+`M$c#_\b}XE0***

Logical Block Corruptions

XC?W%}#h)P0This is when block contains a valid checksum and the structure below the beginning of the block is corrupt (Block content is corrupt). It may cause different ORA-600 errors.
a-x])]"vT*R-VE|0ITPUB个人空间(SM7B8]2X'}~2d i
The detailed corruption description for Logical Corruptions are not normally printed in the alert.log. DBVerify will report what is logically corrupted in the block.
&nu c'i_2]8AI K0
~.p x YQM `F0Corruption Examples are:

• row locked by non-existent transaction - ORA-600 [4512],etc
• the amount of space used is not equal to block size
• avsp bad
• etc.

When db_block_checking is enabled, it may produce the internal errors ORA-600 [kddummy_blkchk] or ORA-600 [kdBlkCheckError].

ITPUB个人空间$NKJyt+z!x%MaM
If db_block_checking is enabled and the block is already logically corrupt on disk, the next block update will mark the block as Soft Corrupt and future reads of this block will produce the error ORA-1578. In that case DBVerify reports this corruption with error "DBV-200: Block, dba , already marked corrupted".

References

NOTE:1088018.1 - Master Note for Handling Oracle Database Corruption Issues
Cn.Q*Gw&jRH)`'DH0 NOTE:1578.1 - OERR: ORA-1578 "ORACLE data block corrupted (file # %s, block # %s)" Master Note ITPUB个人空间j%g`dM
NOTE:28814.1 - Handling Oracle Block Corruptions in Oracle7/8/8i/9i/10g/11g ITPUB个人空间_@*Ru1Lo$oa:S#g
NOTE:794505.1 - ORA-1578 / ORA-26040 Corrupt blocks by NOLOGGING - Error explanation and solution ITPUB个人空间C\2C-NP#Rk6n
NOTE:819533.1 - How to identify the corrupt Object reported by ORA-1578 / RMAN / DBVERIFY
3t&O6H5g;B%O0 NOTE:836658.1 - Identify the corruption extension using RMAN/DBV/ANALYZE etc
8v kF9L0`vv0
1qH5D5AP2F`*D;Bt2b0
n W6t(gl0

一.物理块损坏：

各种各样的块损坏通常是通过ORA-1578报告出来的，详细的信息被记录在alert日志中。物理损坏的例子包括：

>坏头

> 块破坏/不完整 （Fractured Block）

> 块的checksum不正确 (Bad Checksum)

> 块的位置错误 (Block Misplaced)

> 归零的块/ORA-8103

ITPUB个人空间$g:`2c!UCFt9t

二.逻辑块损坏：

当块包含一个正确的checksum和结构，但是块头一下的部分（块的内容）被损坏，可能引起不同的ORA-600错误。

逻辑损坏详细的损坏信息通常不打印在alert日志中，DBV将报告逻辑损坏的块。

逻辑损坏的例子包括：

> 行被不存在的事务锁定-ORA-600[4512]等

> 使用的空间大小不等于块的大小

> avsp bad

> 等等

当启用db_block_checking，可能会产生内部错误ORA-600 [kddummy_blkchk] 或者 ORA-600 [kdBlkCheckError]。

"S1m$Z@%ut0

三.DB_BLOCK_CHECKSUM和DB_BLOCK_CHECKING
tl$q1} o;O0 DB_BLOCK_CHECKSUM用于防止物理IO的损坏，默认值是TYPICAL，在DBWR进程写入磁盘的时候会记录数据的CHECKSUM值，将其存储在块头中，下次在读取的时候会重新计算块的CHECKSUM值，与块头进行比对以判断该块是否损坏。如果将其设置为FULL，还会验证内存中的块的CHECKSUM值，避免内存的问题导致块的损坏。 即使将DB_BLOCK_CHECKSUM值设置为FALSE，对于SYSTEM表空间也会进行相关的验证。 ITPUB个人空间"XBL.fH%[
DB_BLOCK_CHECKING用于逻辑控制块， 如果db_block_checking是启用，磁盘的块已经被逻辑损坏，下一次更新块将以软损坏标记块，将来对这个块的读将产生ORA-1578错误，在这种情况下报告损坏错误"DBV-200:Block,dba ,already marked corrupted"。 即使将DB_BLOCK_CHECKING值设置为FALSE，对于SYSTEM表空间也会进行相关的验证。 ITPUB个人空间/NL{.C&H|L
ITPUB个人空间 i'S7ySeb
四.验证工具介绍 ITPUB个人空间0O^ M%bLj8W0nW j
对于坏块的验证主要有两个工具DBV和RMAN。
+HKs#V&x0 ITPUB个人空间DCx `$JjWr;Eh7^
1.DBV工具
(lW9u.c#i,rDl4U&D#Y(}0 DBV是DBVERIFY的缩写，它是执行物理数据结构完整性检查的外部命令行工具，它能用在脱机或联机的数据库，也可以用在备份文件上。因为DBVERYIFY能对脱机数据库运行，完整性检查速度大幅度提高。DBV检查被限制在缓存管理的块（数据块），DBV只用于数据文件，不能用于控制文件或Redo日志文件的检查。ITPUB个人空间&CRSX#R6PpGwE`0d
dbv help=y可以看到DBV工具参数详细的帮助信息。
#NXvax a0 ITPUB个人空间F5O6vVm+e%w
2.RMAN工具 ITPUB个人空间jN@o1Eao!O/]
下面是有关RMAN的VALIDATE命令验证逻辑坏块和物理坏块的内容： ITPUB个人空间\#g3w^3F1F

How to check for physical and logical database corruption using "backup validate check logical database" command for database on a non-archivelog mode [ID 466875.1]
		修改时间 30-JUN-2011 类型 FAQ 状�?/em> PUBLISHED

In this Document
X an$^d0 Purpose
nQy+K:Lh5NU[0xB0 Questions and Answers
:k `e7WLX/m'bJ S0

---

Purpose

How can we use this command for a non-archivelog database so we can use this as opposed to the dbv command ?
RMAN> backup validate check logical database;

Questions and Answers

"backup validate check logical database" RMAN command could NOT be used for a non-archivelog database.

RMAN> backup validate check logical database;

RMAN-00571: ===========================================================
RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
RMAN-00571: ===========================================================
RMAN-03009: failure of backup command on ORA_DISK_2 channel at 11/06/2007 14:56:31
ORA-19602: cannot backup or copy active file in NOARCHIVELOG mode

The only way to do that for database on no-archive log mode is to perform. above command while the target database in MOUNT status not OPEN

Additional Information:
------------------------------------
- RMAN restore/backup validate with check logical will ensure that there is no logical and also physical corruption.However , DBV would report only physical corruption.

- If the backup validate discovers corrupt blocks, then RMAN updates the V$DATABASE_BLOCK_CORRUPTION view with rows describing the corruptions.

- After corrupt block is repaired the row identifies this block is deleted from the view.

---

从上面的描述可以看出，DBV只报告物理损坏，validate check logical database可以验证物理损坏和逻辑损坏，执行VALIDATE命令后通过查询V$DATABASE_BLOCK_CORRUPTION视图可以获得坏块的详细信息，块被修复后V$DATABASE_BLOCK_CORRUPTION中相应的记录会被删除。该命令只能运行在归档模式下的数据库，如果要在非归档模式下执行该命令，必须使数据库在MOUNT模式。ITPUB个人空间9r2@ A'I1J'?7Xh?
执行下面的命令能验证归档日志的逻辑损坏和物理损坏：
0q5d;{Vq(^5S2r_0RMAN > BACKUP VALIDATE CHECK LOGICAL DATABASE ARCHIVELOG ALL;
MuD&lK0 执行下面的命令验证数据库物理损坏的块：
~%L0_-u7R7lvm0RMAN > BACKUP VALIDATE DATABASE|ARCHIVELOG ALL;ITPUB个人空间]9ew2A"J4i
没有CHECK LOGICAL关键字的BACKUP命令只会验证物理损坏。

来自 “ ITPUB博客 ” ，链接：http://blog.itpub.net/196700/viewspace-738218/，如需转载，请注明出处，否则将追究法律责任。

转载于:http://blog.itpub.net/196700/viewspace-738218/