oracle坏块问题处理
1、说明
坏块问题是经常出现在数据库系统中的,如果没有合适的处理方法往往会导致坏块对象不可用或者数据丢失。本文从坏块产生开始说起,着重说明一但坏块产生后不同情况的处理方法。
2、坏块产生的原因:
硬件的I/O错误
操作系统的I/O错误或缓冲问题
内存或paging问题
磁盘修复工具
一个数据文件的一部分正在被覆盖
Oracle试图访问一个未被格式化的系统块失败
数据文件部分溢出
Oracle或者操作系统的bug
3、坏块的发现:
3.1、Alter日志报错
Tue Aug 17 10:48:07 2010
Corrupt Block Found
TSN = 7, TSNAME = BTEST
RFN 6, BLK = 839, rdba = 25166663
OBJN = 49205, BJD = 49205, BJECT= BOBJ, SUBOBJECT =
Segment wner= AN, Segment Type = Table Segment
其中RFN表示的是relative_fno,6号文件的839块,段类型为表段,是表bobj出现坏块。
3.2、查询报错
SQL> select count(*) from bobj;
select count(*) from bobj
*
ERROR at line 1:
ORA-01578: ORACLE data block corrupted (file # 6, block # 839)
ORA-01110: data file 6: 'F:\ORACLE\PRODUCT\10.1.0\ORADATA\DB10\BTEST.DBF'
3.3、分析表报错
SQL> analyze table bobj validate structure cascade;
analyze table bobj validate structure cascade
*
ERROR at line 1:
ORA-01578: ORACLE data block corrupted (file # 6, block # 839)
ORA-01110: data file 6: 'F:\ORACLE\PRODUCT\10.1.0\ORADATA\DB10\BTEST.DBF'
3.4、Rman备份报错
RMAN> backup tablespace btest;
Starting backup at 17-8月 -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=F:\ORACLE\PRODUCT\10.1.0\ORADATA\DB10\BTEST.DBF
channel ORA_DISK_1: starting piece 1 at 17-8月 -10
RMAN-00571: ===========================================================
RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
RMAN-00571: ===========================================================
RMAN-03009: failure of backup command on ORA_DISK_1 channel at 08/17/2010 11:03:09
ORA-19566: exceeded limit of 0 corrupt blocks for file F:\ORACLE\PRODUCT\10.1.0\ORADATA\DB10\BTEST.DBF
3.5、Dbv检查报错
F:\oracle\product\10.1.0\oradata\db10>dbv file=BTEST.DBF blocksize=8192
DBVERIFY: Release 10.1.0.2.0 - Production on星期二 8月 17 10:49:02 2010
Copyright (c) 1982, 2004, Oracle. All rights reserved.
DBVERIFY - Verification starting : FILE = BTEST.DBF
Page 839 is marked corrupt
Corrupt block relative dba: 0x01800347 (file 6, block 839)
Bad check value found during dbv:
Data in bad block:
type: 6 format: 2 rdba: 0x01800347
last change scn: 0x0000.0005246f seq: 0x1 flg: 0x04
spare1: 0x0 spare2: 0x0 spare3: 0x0
consistency value in tail: 0x246f0601
check value in block header: 0x50c2
computed block checksum: 0x2751
DBVERIFY - Verification complete
Total Pages Examined : 5376
Total Pages Processed (Data) : 5165
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 : 201
Total Pages Marked Corrupt : 1
Total Pages Influx : 0
3.6、查询坏块发生的数据文件以及对应表空间
Select file_name,tablespace_name,file_id “AFN”,relative_fno “RFN”
From dba_data_files;
Select file_name,tablespace_name,file_id, relative_fno“RFN”
From dba_temp_files;
3.7、查询存在坏块的对象是什么:
SELECT tablespace_name, segment_type, owner,
segment_name, partition_name FROM
dba_extents WHERE file_id =v_file_id and v_block_id
between block_id AND block_id + blocks – 1;
4、出现坏块的常见对象:
Sys用户下的对象
回滚段
临时段
索引或者分区索引
表
5、问题的处理
5.1、无备份情况下的恢复
5.1.1、Sys用户下的对象,需要谨慎处理。
5.1.2、回滚段坏块采用类似ora-600(4000)的处理方法将回滚段删除即可。会造成事物失败,数据丢失。
5.1.3、索引:重建,期间会锁表对业务有影响,数据不会丢失。
5.1.4、表:
5.1.4.1、EVENT10231
SQL> select count(*)from bobj;
COUNT(*)
----------
376240
SQL> analyze table bobj validate structure cascade;
analyze table bobj validate structure cascade
*
ERROR at line 1:
ORA-01578: ORACLE data block corrupted (file # 6, block # 495)
ORA-01110: data file 6: 'F:\ORACLE\PRODUCT\10.1.0\ORADATA\DB10\BTEST.DBF'
SQL> alter session set events ='10231 trace name context forever,level 10';
Session altered.
SQL> select count(*) from bobj;
COUNT(*)
----------
376169
忽略了坏块进行读写,所以数据量上有变化,丢失了一部分数据。
5.1.4.2、ROWID RANGE SCAN
FUNCTION ROWID_CREATE RETURNS ROWID
Argument Name Type In/Out Default?
------------------------------ ----------------------- ------ --------
ROWID_TYPE NUMBER IN
OBJECT_NUMBER NUMBER IN
RELATIVE_FNO NUMBER IN
BLOCK_NUMBER NUMBER IN
ROW_NUMBER NUMBER IN
SQL> select * from v$database_block_corruption;
FILE# BLOCK# BLOCKS CORRUPTION_CHANGE# CORRUPTIO
---------- ---------- ---------- ------------------ ---------
6 839 1 0 CHECKSUM
SQL> SELECT dbms_rowid.rowid_create(1,49205,6,839,0) from DUAL;
DBMS_ROWID.ROWID_C
------------------
AAAMA1AAGAAAANHAAA
SQL> SELECT dbms_rowid.rowid_create(1,49205,6,840,0) from DUAL;
DBMS_ROWID.ROWID_C
------------------
AAAMA1AAGAAAANIAAA
SQL> create table bbobj tablespace btest as select * from bobj where 1=2;
Table created.
SQL> insert into bbobj select /*+ rowid(a) */ * from bobj where rowid
AAAANHAAA';
60993 rows created.
SQL> insert into bbobj select /*+ rowid(a) */ * from bobj where rowid>='AAAMA1AAG
AAAANIAAA';
315108 rows created.
SQL> commit;
Commit complete.
SQL> select count(*) from bbobj;
COUNT(*)
----------
376101
从v$database_block_corruption中可以查处当前数据库中的坏块信息
通过调用dbms_rowid.rowid_create确认出坏块对应的rowid,重新创建表结构相同的表,并以rowid为条件跳过坏块将好数据存入到中间表中,损失一部分数据,仍然可以将坏块的影响忽略掉。
5.1.4.3、Dbms_repair
以sys用户执行
PROCEDURE SKIP_CORRUPT_BLOCKS
Argument Name Type In/Out Default?
------------------------------ ----------------------- ------ --------
SCHEMA_NAME VARCHAR2 IN
OBJECT_NAME VARCHAR2 IN
OBJECT_TYPE BINARY_INTEGER IN DEFAULT
FLAGS BINARY_INTEGER IN DEFAULT
SQL> Execute DBMS_REPAIR.SKIP_CORRUPT_BLOCKS('AN','BOBJ')
PL/SQL procedure successfully completed.
SQL> conn an/an
Connected.
SQL> select count(*) from bobj;
COUNT(*)
----------
376169
试用sys用户调用dbms_repair中的skip_corrupt_blocks忽略对象中的坏块,同样会丢失坏块中的数据。包中还包含过程FIX_CORRUPT_BLOCKS,可以fix掉坏块。
5.2、有备份情况下的恢复
5.2.1、Blockrecover
RMAN> blockrecover datafile 5 block 425;
Starting blockrecover at 17-8月-10
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 00005
channel ORA_DISK_1: restored block(s) from backup piece 1
piece handle=F:\ORACLE\PRODUCT\10.1.0\FLASH_RECOVERY_AREA\DB10\BACKUPSET\2010_08_17\O1_MF_NNNDF_TAG20100817T162508_66NKV5B8_.BKP tag=TAG20100817T162508
channel ORA_DISK_1: block restore complete
starting media recovery
media recovery complete
Finished blockrecover at 17-8月-10
SQL> select count(*) from an.atest;
COUNT(*)
----------
376240
5.2.2、Recover datafile
SQL> select count(*) from an.atest;
select count(*) from an.atest
*
ERROR at line 1:
ORA-01578: ORACLE data block corrupted (file # 5, block # 963)
ORA-01110: data file 5: 'F:\ORACLE\PRODUCT\10.1.0\ORADATA\DB10\ATEST.DBF'
RMAN> sql " alter database datafile 5 offline";
sql statement: alter database datafile 5 offline
RMAN> restore datafile 5;
Starting restore at 17-8月-10
using channel ORA_DISK_1
channel ORA_DISK_1: starting datafile backupset restore
channel ORA_DISK_1: specifying datafile(s) to restore from backup set
restoring datafile 00005 to F:\ORACLE\PRODUCT\10.1.0\ORADATA\DB10\ATEST.DBF
channel ORA_DISK_1: restored backup piece 1
piece handle=F:\ORACLE\PRODUCT\10.1.0\FLASH_RECOVERY_AREA\DB10\BACKUPSET\2010_08_17\O1_MF_NNNDF_TAG20100817T162508_66NKV5B8_.BKP tag=TAG20100817T162508
channel ORA_DISK_1: restore complete
Finished restore at 17-8月-10
RMAN> recover datafile 5;
Starting recover at 17-8月-10
using channel ORA_DISK_1
starting media recovery
media recovery complete
Finished recover at 17-8月-10
RMAN> sql " alter database datafile 5 online";
sql statement: alter database datafile 5 online
SQL> select count(*) from an.atest;
COUNT(*)
----------
376240
5.3、Rman备份跳过坏块
SET MAXCORRUPT FOR DATAFILE filename TO n;
例:
run {
allocate channel node1 type disk;
SET MAXCORRUPT FOR DATAFILE 8 TO 3;
set limit channel node1 kbytes = 1800000;
backup as compressed backupset full database format '$BACKUPDIR/full_%d_%T_%s_%p' plus archivelog format '$BACKUPDIR/arch_%d_%T_%s_%p' delete all input;
backup current controlfile format '$BACKUPDIR/ctl_%d_%T_%s_%p' TAG "control.bak";
release channel node1;
}