如何使用 RMAN 识别数据库中损坏的对象 (文档 ID 1623348.1)

文档内容


目标

解决方案
  步骤1:识别坏块
  步骤2:识别损坏的段

参考


适用于:

Oracle Database - Enterprise Edition - 版本 8.1.7.0 到 11.2.0.4 [发行版 8.1.7 到 11.2]
本文档所含信息适用于所有平台

目标

如何使用 RMAN 识别数据库中损坏的段。

解决方案

步骤1:识别坏块

执行下面的 RMAN 命令,使所有的坏块信息被记录在 v$database_block_corruption 视图中:

RMAN> backup validate check logical database;


注意:

这个命令只是检查数据库的坏块,而不会真正进行备份。从 11g 开始可以省略 backup 子句,而直接使用命令"validate check logical database"。

如果由于缺失文件导致命令失败,可以增加 'SKIP INACCESSIBLE' 子句来避免这个问题。

为了加快检查速度,可以设置 PARALLELISM 指定多个通道:

 

RMAN> configure device type disk parallelism 4;
RMAN> backup validate check logical database;

OR

RMAN> run {
allocate channel d1 type disk;
allocate channel d2 type disk;
allocate channel d3 type disk;
allocate channel d4 type disk;
backup validate check logical database;
}

输出

坏块信息会被记录在视图 V$DATABASE_BLOCK_CORRUPTION 中。11g RMAN 会生成一个 trace 文件,详细描述坏块信息:

RMAN VALIDATE 屏幕输出:

File Status Marked Corrupt Empty Blocks Blocks Examined High SCN
---- ------ -------------- ------------ --------------- ----------
6    FAILED 0              501          640             1950088   
  File Name: /oracle/dbs/users.dbf
  Block Type Blocks Failing Blocks Processed
  ---------- -------------- ----------------
  Data       9              9              
  Index      0              0               
  Other      0              130             

validate found one or more corrupt blocks
See trace file /oracle/log/diag/rdbms/orcl/orcl/trace/orcl_ora_28424.trc for details
Finished validate at

Trace 文件输出坏块信息,这个例子描述了 2 个坏块,一个物理坏块(file 6 block 9)和一个逻辑坏块(file 6 block 10):

Corrupt block relative dba: 0x01000009 (file 4, block 9)
Bad check value found during validation
Data in bad block:
 type: 16 format: 2 rdba: 0x01000009
 last change scn: 0x0000.00000000 seq: 0xff flg: 0x04
 spare1: 0x0 spare2: 0x0 spare3: 0x0
 consistency value in tail: 0x000010ff
 check value in block header: 0xb4e0
 computed block checksum: 0xa800
Reread of blocknum=9, file=/oracle/dbs/users.dbf found same corrupt data

Block Checking: DBA = 25165834, Block Type = KTB-managed data block
data header at 0x2b2deb49e07c
kdbchk: fsbo(144) wrong, (hsz 78)
Error backing up file 6, block 10: logical corruption

坏块信息记录在视图 V$DATABASE_BLOCK_CORRUPTION 中:

SQL> select * from V$DATABASE_BLOCK_CORRUPTION;

          FILE#          BLOCK#          BLOCKS CORRUPTION_CHANGE# CORRUPTIO
--------------- --------------- --------------- ------------------ ---------
              6              10               1      8183236781662 LOGICAL
              6              42               1                  0 FRACTURED
              6              34               2                  0 CHECKSUM
              6              50               1      8183236781952 LOGICAL
              6              26               4                  0 FRACTURED

5 rows selected.

注意:

  • CHECK LOGICAL 选项既会检查物理坏块也会检查逻辑坏块。坏块类型解释参见 Note 840978.1
  • 当发现逻辑坏块,alert 日志会更新以下的信息:

    Error backing up file , block : logical corruption

    在 11g 会生成一个 trace 文件,描述坏块信息。
  • 当发现物理坏块时,alert 日志也会更新下面的信息:
Corrupt block relative dba: 0x01000009 (file 4, block 9)
Bad check value found during validation
Data in bad block:
 type: 16 format: 2 rdba: 0x01000009
 last change scn: 0x0000.00000000 seq: 0xff flg: 0x04
 spare1: 0x0 spare2: 0x0 spare3: 0x0
 consistency value in tail: 0x000010ff
 check value in block header: 0xb4e0
 computed block checksum: 0xa800
Reread of blocknum=9, file=/oracle/dbs/users.dbf found same corrupt data
  • 检查单独一个数据文件或者特定的数据文件,使用命令"check logical validate datafile 1, 2"。
  • 监控 VALIDATE 命令运行的进度,请运行下面的查询:
select sid, serial#, context, sofar, totalwork,round(sofar/totalwork*100,2) "%_complete"
from v$session_longops
where opname like 'RMAN%'
  and opname not like '%aggregate%'
  and totalwork != 0
  and sofar <> totalwork;
  • 10g 或者之前版本的 NOARCHIVELOG 数据库,数据库必须在 MOUNT 状态下运行 VALIDATE 命令,否则报 ORA-19602,如果不能关闭数据库,使用 dbverify 代替 RMAN。这个限制在 11g 取消。
  • 11g或者更高版本:可以使用 section 子句并行校验一个单独的数据文件。
    RMAN 将数据文件分割为多个部分,并行处理每个部分。下面的例子中数据文件 5 被分割为 1gb 大小的部分,如果配置或者分配了多个通道,那么 RMAN 会并行校验每个部分。(参考: The Oracle Database Backup and Recovery User's Guide for more information):
backup validate check logical datafile 5 SECTION SIZE 1024M;
  • 11g 或者更高版本可以使用 BLOCK TO 子句检查一定范围内的数据块。下面命令检查数据文件 1 中的数据块 5 到 20:
validate check logical datafile 1 BLOCK 5 TO 20;
  • 在 8i 版本中,如果 RMAN validate 命令检查发现坏块,那么只会记录在 alert 日志中。用户必须在alert日志中查看 validate 命令运行时间范围内的日志。坏块信息不会返回到 RMAN 接口。在 9i 及以上版本中,如果 RMAN validate 命令检查发现坏块,那么坏块信息可以通过查询视图 V$DATABASE_BLOCK_CORRUPTION 得到。
  • 每次运行 RMAN backup validate 命令,视图 V$DATABASE_BLOCK_CORRUPTION 中坏块信息都会被更新。了解视图内容,参见 Oracle? Database Reference 文档。

步骤2:识别损坏的段

下面的查询可以将视图 v$database_block_corruption 中记录的坏块匹配到对应的段或者是空闲块。

$ sqlplus / as sysdba
set pagesize 2000
set linesize 280
SELECT e.owner, e.segment_type, e.segment_name, e.partition_name, c.file#
     , greatest(e.block_id, c.block#) corr_start_block#
     , least(e.block_id+e.blocks-1, c.block#+c.blocks-1) corr_end_block#
     , least(e.block_id+e.blocks-1, c.block#+c.blocks-1)
       - greatest(e.block_id, c.block#) + 1 blocks_corrupted
     , corruption_type description
  FROM dba_extents e, v$database_block_corruption c
 WHERE e.file_id = c.file#
   AND e.block_id <= c.block# + c.blocks - 1
   AND e.block_id + e.blocks - 1 >= c.block#
UNION
SELECT s.owner, s.segment_type, s.segment_name, s.partition_name, c.file#
     , header_block corr_start_block#
     , header_block corr_end_block#
     , 1 blocks_corrupted
     , corruption_type||' Segment Header' description
  FROM dba_segments s, v$database_block_corruption c
 WHERE s.header_file = c.file#
   AND s.header_block between c.block# and c.block# + c.blocks - 1
UNION
SELECT null owner, null segment_type, null segment_name, null partition_name, c.file#
     , greatest(f.block_id, c.block#) corr_start_block#
     , least(f.block_id+f.blocks-1, c.block#+c.blocks-1) corr_end_block#
     , least(f.block_id+f.blocks-1, c.block#+c.blocks-1)
       - greatest(f.block_id, c.block#) + 1 blocks_corrupted
     , 'Free Block' description
  FROM dba_free_space f, v$database_block_corruption c
 WHERE f.file_id = c.file#
   AND f.block_id <= c.block# + c.blocks - 1
   AND f.block_id + f.blocks - 1 >= c.block#
order by file#, corr_start_block#;



输出例子:

OWNER SEGMENT_TYPE       SEGMENT_NAME PARTITION_ FILE# CORR_START_BLOCK# CORR_END_BLOCK# BLOCKS_CORRUPTED DESCRIPTION
----- ------------------ ------------ ---------- ----- ----------------- --------------- ---------------- -------------
SCOTT TABLE              EMP                         6                10              10                1
SCOTT TABLE PARTITION    ORDER        ORDER_JAN      6                26              28                3
                                                     6                29              29                1 Free Block
SCOTT TABLE              BONUS                       6                34              34                1
                                                     6                35              35                1 Free Block
SCOTT TABLE              DEPT                        6                42              42                1 Segment Header
SCOTT TABLE              INVOICE                     6                50              50                1


注意:

  • 如果坏块出现在字典管理的表空间,并且段头块损坏,上面的查询可能显示同一个块 2 次。
  • 如果 ASSM 表空间的段头块损坏,那么上面的查询会显示该段头块,但是相同对象上的其他坏块可能就不会再显示。

识别出12C中NOLOGGING的坏块
通过RMAN中的新视图v$nonlogged_block来识别12C中NOLOGGING块

$ sqlplus / as sysdba

set echo on

select systimestamp
from dual;

select FILE#, BLOCK#, BLOCKS, to_char(NONLOGGED_START_CHANGE#, '999999999999999') NONLOGGED_START_CHANGE#
from v$nonlogged_block;

set pagesize 2000
set linesize 250

SELECT e.owner, e.segment_type, e.segment_name, e.partition_name, c.file#
, greatest(e.block_id, c.block#) corr_start_block#
, least(e.block_id+e.blocks-1, c.block#+c.blocks-1) corr_end_block#
, least(e.block_id+e.blocks-1, c.block#+c.blocks-1)
- greatest(e.block_id, c.block#) + 1 blocks_corrupted
, null description
FROM dba_extents e, v$nonlogged_block c
WHERE e.file_id = c.file#
AND e.block_id <= c.block# + c.blocks - 1
AND e.block_id + e.blocks - 1 >= c.block#
UNION
SELECT null owner, null segment_type, null segment_name, null partition_name, c.file#
, greatest(f.block_id, c.block#) corr_start_block#
, least(f.block_id+f.blocks-1, c.block#+c.blocks-1) corr_end_block#
, least(f.block_id+f.blocks-1, c.block#+c.blocks-1)
- greatest(f.block_id, c.block#) + 1 blocks_corrupted
, 'Free Block' description
FROM dba_free_space f, v$nonlogged_block c
WHERE f.file_id = c.file#
AND f.block_id <= c.block# + c.blocks - 1
AND f.block_id + f.blocks - 1 >= c.block#
order by file#, corr_start_block#;

 

 

参考

NOTE:377146.1 - How to Check Archivelogs for Corruption using RMAN
NOTE:836658.1 - Identify the Corruption Extension for Block Corruption, Table/Index Inconsistency, Data Dictionary and Lost Writes
NOTE:561010.1 - Which Blocks Will RMAN Check For Corruption Or Include In A Backupset?
NOTE:819533.1 - How to identify the corrupt Object reported by ORA-1578 / RMAN / DBVERIFY
NOTE:144911.1 - RMAN : Block-Level Media Recovery - Concept & Example
NOTE:840978.1 - Physical and Logical Block Corruptions. All you wanted to know about it.
NOTE:338607.1 - How To Check (Validate) If RMAN Backup(s) Are Good
NOTE:144911.1 - RMAN : Block-Level Media Recovery - Concept & Example
NOTE:472231.1 - How to identify all the Corrupted Objects in the Database with RMAN
NOTE:794505.1 - ORA-1578 / ORA-26040 Corrupt blocks by NOLOGGING - Error explanation and solution
请使用浏览器的分享功能分享到微信等