10种 Oracle 文件损坏及恢复的过程

一、 数据库服务器基本情况
    OS：RHEL 3

    CPU：4个Intel(R) Xeon(TM) MP CPU 2.70GHz
    Mem：8G
    Swap：16G
    Disk：120G

    Oracle Database 10g Enterprise Edition Release 10.1.0.3.0

二、 备份方式
    数据库以archive模式运行，RMAN多级增量备份。策略如下：
    设置控制文件自动备份。
    每三个月做一个数据库的全备份（包括所有得数据库和只读表空间）,并备份归档日志。
    每一个月做一次零级备份（不包含只读表空间），并备份归档日志。
    每周做一次一级备份，并备份归档日志。
    每天做一次二级备份，并备份归档日志。

三、 恢复案例
    所有恢复的前提：已经做过数据库全备份（包括归档日志），控制文件和spfile自动备份。

    1． 损坏一个数据文件
    （1）故障模拟
        删除数据文件：rm /u02/oradata/dbnms/users01.dbf
        关闭数据库：shutdown immediate;
        ORA-01116: error in opening database file 4
        ORA-01110: data file 4: '/u02/oradata/dbnms/users01.dbf'
        ORA-27041: unable to open file
        Linux Error: 2: No such file or directory
        Additional information: 3
        强行关闭：sutdown abort;
        启动数据库：startup;
        ORA-01157: cannot identify/lock data file 4 - see DBWR trace file
        ORA-01110: data file 4: '/u02/oradata/dbnms/users01.dbf'
    （2）恢复步骤
        rman target sys@dbnms catalog rmanuser@cata
        run{
            allocate channel c1 type disk;
            restore datafile 4;
            recover datafile 4;
            sql 'alter database datafile 4 online';
            sql 'alter database open';
            release channel c1;
            }

        sqlplus sys as sysdba
        select instance_name,status from v$instance;
        INSTANCE_NAME STATUS
        ---------------- ------------
        dbnms OPEN

    恢复成功

    2． 损坏全部数据文件
    （1）故障模拟
        删除数据文件：rm /u02/oradata/dbnms/*.dbf
        强行关闭：sutdown abort;
        启动数据库：startup;
        ORA-01157: cannot identify/lock data file 1 - see DBWR trace file
        ORA-01110: data file 1: '/u02/oradata/dbnms/system01.dbf'

    （2）恢复步骤
        rman target sys@dbnms catalog rmanuser@cata
        run{
            allocate channel c1 type disk;
            restore database;
            recover database;
            sql 'alter database open';
            release channel c1;
            }

        sqlplus sys as sysdba
        select instance_name,status from v$instance;
        INSTANCE_NAME STATUS
        ---------------- ------------
        dbnms OPEN

        恢复临时文件：
        alter database tempfile '/u02/oradata/dbnms/temp01.dbf' drop;
        alter tablespace temp add tempfile '/u02/oradata/dbnms/temp01.dbf' size 50M autoextend on next 5M maxsize unlimited;

    恢复成功

    3． 损坏非当前联机日志成员
    （1）故障模拟
         删除日志文件：rm /u02/oradata/dbnms/redo01.log
        关闭数据库：shutdown immediate;
        启动数据库：startup;
        select * from v$logfile;

        GROUP# STATUS TYPE MEMBER IS_
        ---------- ------- ------- ------------------------------ ---
        3 ONLINE /u02/oradata/dbnms/redo03.log NO
        2 STALE ONLINE /u02/oradata/dbnms/redo02.log NO
        1 INVALID ONLINE /u02/oradata/dbnms/redo01.log NO
        1 STALE ONLINE /u02/oradata/dbnms/redo11.log NO
        1 STALE ONLINE /u02/oradata/dbnms/redo21.log NO
        2 STALE ONLINE /u02/oradata/dbnms/redo12.log NO
        3 ONLINE /u02/oradata/dbnms/redo13.log NO
        2 STALE ONLINE /u02/oradata/dbnms/redo22.log NO
        3 ONLINE /u02/oradata/dbnms/redo23.log NO
        4 ONLINE /u02/oradata/dbnms/redo31.log NO
        4 ONLINE /u02/oradata/dbnms/redo32.log NO

        GROUP# STATUS TYPE MEMBER IS_
        ---------- ------- ------- ------------------------------ ---
        4 ONLINE /u02/oradata/dbnms/redo33.log NO

    （2）恢复步骤
        alter database drop logfile member '/u02/oradata/dbnms/redo01.log';
        alter database add logfile member '/u02/oradata/dbnms/redo01.log' to group 1;

    恢复成功

    4． 损坏非当前联机日志组
    （1）故障模拟
        删除日志文件组1的所有文件：
        rm /u02/oradata/dbnms/redo01.log
        rm /u02/oradata/dbnms/redo11.log
        rm /u02/oradata/dbnms/redo21.log
        关闭数据库：shutdown immediate;
        启动数据库：startup;
        Database mounted.
        ORA-00313: open failed for members of log group 1 of thread 1
        ORA-00312: online log 1 thread 1: '/u02/oradata/dbnms/redo11.log'
        ORA-00312: online log 1 thread 1: '/u02/oradata/dbnms/redo21.log'
        ORA-00312: online log 1 thread 1: '/u02/oradata/dbnms/redo01.log'

    （2）恢复步骤
        alter database clear logfile group 1;
        如果该日志组还没有归档，则用：
        alter database clear unarchived logfile group 1;
        打开数据库：
        alter database open;

    恢复成功

    5． 损坏全部联机日志
    （1）故障模拟
        删除日志文件：rm /u02/oradata/dbnms/*.log
        关闭数据库：shutdown immediate;
        启动数据库：startup;
        Database mounted.
        ORA-00313: open failed for members of log group 1 of thread 1
        ORA-00312: online log 1 thread 1: '/u02/oradata/dbnms/redo11.log'
        ORA-00312: online log 1 thread 1: '/u02/oradata/dbnms/redo21.log'
        ORA-00312: online log 1 thread 1: '/u02/oradata/dbnms/redo01.log'

    （2）恢复步骤
        shutdown immediate;
        create pfile from spfile;
        cd $ORACLE_HOME/dbs
        vi initdbnms.ora
        加一个参数：_allow_resetlogs_corruption=true
        create spfile from pfile;
        startup;
        Database mounted.
        ORA-00313: open failed for members of log group 1 of thread 1
        ORA-00312: online log 1 thread 1: '/u02/oradata/dbnms/redo11.log'
        ORA-00312: online log 1 thread 1: '/u02/oradata/dbnms/redo21.log'
        ORA-00312: online log 1 thread 1: '/u02/oradata/dbnms/redo01.log'
        recover database until cancel;
        alter database open resetlogs;
        同步catalog：
        rman target sys@dbnms catalog rmanuser@cata
        reset database;

        new incarnation of database registered in recovery catalog
        starting full resync of recovery catalog
        full resync complete

        做一个full备份：
        /home/oracle/dbbat/backup_full.sh

    恢复成功

    6． 损坏一个控制文件
    （1）故障模拟
        删除控制文件：rm /u02/oradata/dbnms/control01.ctl
        关闭数据库：shutdown immediate;
        ORA-00210: cannot open the specified controlfile
        ORA-00202: controlfile: '/u02/oradata/dbnms/control01.ctl'
        ORA-27041: unable to open file
        Linux Error: 2: No such file or directory
        Additional information: 3

    （2）恢复步骤
        拷贝一个好的控制文件：
        cp control02.ctl control01.ctl
        shutdown immediate;
        startup;
        Database mounted.
        ORA-01122: database file 1 failed verification check
        ORA-01110: data file 1: '/u02/oradata/dbnms/system01.dbf'
        ORA-01207: file is more recent than controlfile - old controlfile

        rman target sys@dbnms catalog rmanuser@cata
        run{
            allocate channel c1 type disk;
            restore database;
            recover database;
            sql 'alter database open';
            release channel c1;
            }
        RMAN-00571: ===========================================================
        RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
        RMAN-00571: ===========================================================
        RMAN-03009: failure of sql command on default channel at 09/08/2005 17:29:04
        RMAN-11003: failure during parse/execution of SQL statement: alter database open
        ORA-00322: log 3 of thread 1 is not current copy
        ORA-00312: online log 3 thread 1: '/u02/oradata/dbnms/redo03.log'
        ORA-00312: online log 3 thread 1: '/u02/oradata/dbnms/redo13.log'
        ORA-00312: online log 3 thread 1: '/u02/oradata/dbnms/redo23.log'

        alter database clear unarchived logfile group 3;
        alter database open;

    恢复成功

    7． 损坏全部控制文件
    （1）故障模拟
        删除控制文件：rm /u02/oradata/dbnms/control01.ctl
        关闭数据库：shutdown immediate;
        ORA-00210: cannot open the specified controlfile
        ORA-00202: controlfile: '/u02/oradata/dbnms/control01.ctl'
        ORA-27041: unable to open file
        Linux Error: 2: No such file or directory
        Additional information: 3

    （2）恢复步骤
        shutdown abort;
        startup nomount;

        rman target sys catalog rmanuser@cata
        run{
            allocate channel c1 type disk;
            restore controlfile;
            restore database;
            sql 'alter database mount';
            recover database;
            sql 'alter database open resetlogs';
            release channel c1;
            }

        RMAN-00571: ===========================================================
        RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
        RMAN-00571: ===========================================================
        RMAN-03002: failure of recover command at 09/08/2005 17:43:31
        RMAN-06054: media recovery requesting unknown log: thread 1 seq 9 lowscn 670233

        alter database open resetlogs;

        做一个full备份：
        /home/oracle/dbbat/backup_full.sh

    恢复成功

    8． 损坏临时数据文件
    （1）故障模拟
        删除临时数据文件：rm /u02/oradata/dbnms/temp01.dbf
        关闭数据库：shutdown immediate;
        启动数据库：startup;

    （2）恢复步骤
        alter database tempfile '/u02/oradata/dbnms/temp01.dbf' drop;
        ERROR at line 1:
        ORA-01516: nonexistent log file, datafile, or tempfile
        "/u02/oradata/dbnms/temp01.dbf"

        alter tablespace temp add tempfile '/u02/oradata/dbnms/temp01.dbf' size 50M autoextend on next 5M maxsize unlimited;

    恢复成功

    9． 损坏spfile参数文件
    （1）故障模拟
        删除spfile文件：rm $ORACLE_HOME/dbs/spfiledbnms.ora
        关闭数据库：shutdown immediate;
        启动数据库：startup;

    （2）恢复步骤
        startup nomount;

        rman target sys catalog rmanuser@cata
        restore spfile;

        shutdown immediate;
        startup;

    恢复成功

    10． 损坏全部文件（包括全部数据文件、控制文件、临时数据文件、联机日志文件）
    （1）故障模拟
        删除全部文件：rm /u02/oradata/dbnms/*
        关闭数据库：shutdown immediate;
        ORA-03113: end-of-file on communication channel

    （2）恢复步骤
        sqlplus sys as sysdba
        startup nomount;

        rman target sys catalog rmanuser@cata
        还原控制文件：restore controlfile;
        还原数据库：restore database;
        mount 数据库：alter database mount;
        恢复数据库：recover database;
        RMAN-00571: ===========================================================
        RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
        RMAN-00571: ===========================================================
        RMAN-03002: failure of recover command at 09/09/2005 10:33:13
        RMAN-06054: media recovery requesting unknown log: thread 1 seq 19 lowscn 718284
        用resetlogs方式打开数据库：alter database open resetlogs;
        database opened
        new incarnation of database registered in recovery catalog
        starting full resync of recovery catalog
        full resync complete
        重建临时文件：
        sqlplus sys as sysdba
        alter database tempfile '/u02/oradata/dbnms/temp01.dbf' drop;
        ERROR at line 1:
        ORA-01516: nonexistent log file, datafile, or tempfile
        "/u02/oradata/dbnms/temp01.dbf"

        alter tablespace temp add tempfile '/u02/oradata/dbnms/temp01.dbf' size 50M autoextend on next 5M maxsize unlimited;

        执行一次全库备份：/home/oracle/dbbat/backup_full.sh

    恢复成功