===================================================
如何从磁带中恢复 备份是制定了格式的控制文件??????????
备份时候采用
CONFIGURE CONTROLFILE AUTOBACKUP ON;
CONFIGURE CONTROLFILE AUTOBACKUP FORMAT FOR DEVICE TYPE 'SBT_TAPE' TO 'contr_%F.bk';
在恢复的时候 不知道该如何恢复
====================================================
一次基于时间的oracle不完全恢复（一）
今天，用户让我去解决TSM管理员密码过期的问题，结果过去后，又说他们的oracle数据库的一个表在上周被开发删除了，现在又想要找回来。本来以为很简单的事情，结果搞了一天，直到晚上23点才弄完。
一、方法
由于没有做exp逻辑备份，而且不能停客户的应用；所有只有靠RMAN备份的基于时间点的不完全恢复来做，大体方法如下：找一台和当前环境差不多的主机，做全库恢复，然后recover到想要的时间点，再从这个临时恢复出来的库中exp出用户想要的表，再imp到当前的生产库中去。环境是两台570用HACMP做互为备份的集群，安装AIX5.3,oracle9206，采用TSM5.3做备份。刚好可以使用另外一台主机来做恢复。
顺便说一句，TSM的node在mannger上注册的时候与主机名和IP都没有关系的，所以可以很轻松的实现异机恢复，我想这应该算是TSM最大的优点吧：）。
二、过程
用户的两台570的主机名分别是p570a和p570b，目前oracle跑在p570a上，数据库使用的是裸设备。为了简单起见，在p570b上恢复成文件系统方式。以下是详细的恢复过程：
1、 准备环境：
在p570a上，copy出最新的contorlfile和pfile：
p570a&oracle> rman target / catalog rman/rman@rman
Recovery Manager: Release 9.2.0.6.0 - 64bit Production
Copyright (c) 1995, 2002, Oracle Corporation. All rights reserved.
connected to target database: ORACLE9I (DBID=3231295384)
connected to recovery catalog database
RMAN> run{
2> ALLOCATE CHANNEL ch00 TYPE DISK;
3> copy current controlfile to '/oracle/ctlbk.ctl';
4> RELEASE CHANNEL ch00;
5> }
allocated channel: ch00
channel ch00: sid=133 devtype=DISK
Starting copy at 2006-06-20:10:46:26
channel ch00: copied current controlfile
output filename=/oracle/ctlbk.ctl
Finished copy at 2006-06-20:10:46:27
Starting Control File and SPFILE Autobackup at 2006-06-20:10:46:27
piece handle=/oracle/app/oracle/product/9.2.0.6/dbs/c-3231295384-20060620-04 comment=NONE
Finished Control File and SPFILE Autobackup at 2006-06-20:10:46:30
released channel: ch00
RMAN> quit
Recovery Manager complete.
通过spfile创建pfile：
p570a&oracle> sqlplus "/as sysdba"
SQL*Plus: Release 9.2.0.6.0 - Production on Tue Jun 20 10:46:39 2006
Copyright (c) 1982, 2002, Oracle Corporation. All rights reserved.
Connected to:
Oracle9i Enterprise Edition Release 9.2.0.6.0 - 64bit Production
With the Partitioning and OLAP options
JServer Release 9.2.0.6.0 - Production
SQL> show parameter spfile;
NAME TYPE VALUE
------------------------------------ ----------- ------------------------------
spfile string /dev/rspfile_rw
SQL> create pfile='/oracle/ora9ipfile.ora' from spfile='/dev/rspfile_rw';
File created.
SQL> quit
在p570b上，本来需要安装和a机相同的oracle代码（当前是安装在共享盘上，随HACMP切换使用的，现在也觉得oracle装在internal disk上还是有好处的：）），不过在这个项目做完的时候我做过类似的测试，找找看还在不在：
p570b# varyonvg data04vg
p570b# lsvg -l data04vg
data04vg:
LV NAME TYPE LPs PPs PVs LV STATE MOUNT POINT
test_lv jfs2 1280 1280 1 closed/syncd /test
loglv02 jfs2log 1 1 1 closed/syncd N/A
还好还在：），这个部分的oracle代码是直接从a机复制过来的，所以应该一致：
p570b# mount /test
p570b# cd /test
p570b# ls -l
总计 184
drwxr-xr-x 2 root system 4096 8月23 2005 TT_DB
drwxr-xr-x 3 oracle dba 256 8月23 2005 app
drwxr-xr-x 2 oracle dba 4096 9月27 2005 arch
drwxr-xr-x 2 oracle dba 256 8月22 2005 lost+found
drwxr-xr-x 4 oracle dba 256 9月01 2005 tsm
-rw-r--r-- 1 root system 85797 9月07 2005 tsmconf.tar.Z
呵呵，还在，直接拿过来使用就可以了，不用再倒腾oracle代码了。注意为了不破坏系统的正常状态，所以我把oracle代码放在了test文件系统下，但是oracle的home是在/oracle下面的，于是需要做一个link：
p570b&oracle> cd /oracle
p570b&oracle> ls -l
总计 0
lrwxrwxrwx 1 root system 9 9月27 2005 app -> /test/app
lrwxrwxrwx 1 root system 9 9月27 2005 tsm -> /test/tsm
ok，可以把controlfile和spfile复制过来了：
p570b&oracle> rcp p570a:/oracle/ctlbk.ctl ./
p570b&oracle> ls -l
总计 4568
-rw-r----- 1 oracle dba 2334720 6月20 10时47 ctlbk.ctl
p570b&oracle> rcp p570a:/oracle/ora9ipfile.ora ./
p570b&oracle> ls
ctlbk.ctl ora9ipfile.ora
把pfile复制到oracle的默认位置：
p570b&oracle> cp /oracle/app/oracle/oradata/ora9i/*.ora /oracle/app/oracle/product/9.2.0.6/dbs/
修改pfile参数，主要是controlfile、arch dest和sort_area_size：
p570b&oracle> more initora9i.ora
*.aq_tm_processes=1
*.background_dump_dest='/oracle/app/oracle/admin/oracle9i/bdump'
*.compatible='9.2.0.0.0'
*.control_files='/oracle/app/oracle/oradata/oracle9i/contorl1.ctl'
*.core_dump_dest='/oracle/app/oracle/admin/oracle9i/cdump'
*.db_block_size=8192
*.db_cache_size=209715200
*.db_domain=''
*.db_file_multiblock_read_count=16
*.db_name='oracle9i'
*.dispatchers='(PROTOCOL=TCP)'
*.fast_start_mttr_target=300
*.hash_join_enabled=TRUE
*.instance_name='ora9i'
*.java_pool_size=1048576000
*.job_queue_processes=10
*.large_pool_size=83886080
*.log_archive_dest='/test/arch'
*.log_archive_format='arch%S.arc'
*.log_archive_max_processes=2
*.log_archive_start=true
*.open_cursors=300
*.pga_aggregate_target=209715200
*.processes=150
*.query_rewrite_enabled='FALSE'
*.remote_login_passwordfile='EXCLUSIVE'
*.shared_pool_size=1048576000
*.sort_area_size=20000000
*.star_transformation_enabled='FALSE'
*.timed_statistics=TRUE
*.undo_management='AUTO'
*.undo_retention=10800
*.undo_tablespace='UNDOTBS1'
*.user_dump_dest='/oracle/app/oracle/admin/oracle9i/udump'
p570b&oracle> mv ora9ipfile.ora initora9i.ora
p570b&oracle> sqlplus "/as sysdba"
好，启动oracle到mount状态，开始干活：
SQL*Plus: Release 9.2.0.6.0 - Production on Tue Jun 20 10:53:05 2006
Copyright (c) 1982, 2002, Oracle Corporation. All rights reserved.
Connected to an idle instance.
SQL> startup mount;
ORACLE instance started.
Total System Global Area 2451542600 bytes
Fixed Size 746056 bytes
Variable Size 2231369728 bytes
Database Buffers 218103808 bytes
Redo Buffers 1323008 bytes
ORA-00205: error in identifying controlfile, check alert log for more info
SQL> show parameter controlfile;
SQL> show parameter cont;
NAME TYPE
------------------------------------ ----------------------
VALUE
------------------------------
control_file_record_keep_time integer
7
control_files string
/oracle/app/oracle/oradata/ora
9i/control01.dbf
global_context_pool_size string
SQL> quit
Disconnected from Oracle9i Enterprise Edition Release 9.2.0.6.0 - 64bit Production
With the Partitioning and OLAP options
JServer Release 9.2.0.6.0 - Production
p570b&oracle> ls -l /test/app/oracle/oradata/ora9i
总计 4576
-rw-r----- 1 oracle dba 2334720 6月20 10时47 crontrol01.dbf
-rw-r--r-- 1 oracle dba 1057 6月20 10时49 ora9ipfile.ora
先查以下file#和file_name的对应关系，等下会用到：
SQL> select file#,name from v$datafile;
FILE#
----------
NAME
--------------------------------------------------------------------------------
1
/dev/rsystem_rw
2
/dev/rundo1_rw
3
/dev/rcwmlite_rw
FILE#
----------
NAME
--------------------------------------------------------------------------------
4
/dev/rdrsys_rw
5
/dev/rdata4_1_01_rw
6
/dev/rtools_rw
FILE#
----------
NAME
--------------------------------------------------------------------------------
7
/dev/rusers_rw
8
/dev/rxdb_rw
9
/dev/rdata10_1_01_rw
FILE#
----------
NAME
--------------------------------------------------------------------------------
10
/dev/rdata10_1_02_rw
11
/oracle/app/oracle/product/9.2.0.6/dbs/spaces.dbf
12
/dev/rdata1_1_01_rw
SQL> quit
2、从磁带库中取出需要的备份文件及archivelog
OK，先resotre database，注意，只是restore数据文件，使用until time方法来做，oracle会根据catalog的信息找到离我们需要恢复的时间点最近的上一个全备份，除了time，还有基于SCN、logseq等的方法，后面我们会用到：
p570b&oracle> rman target / catalog rman/rman@rman
Recovery Manager: Release 9.2.0.6.0 - 64bit Production
Copyright (c) 1995, 2002, Oracle Corporation. All rights reserved.
connected to target database: ORACLE9I (DBID=3231295384)
connected to recovery catalog database
RMAN> run {
2> set archivelog destination to '/test/arch/';
3> set until time "to_date('06/13/06 13:00:00','MM/DD/YY HH24:MI:SS')" ;
4> allocate channel t1 type 'sbt_tape' parms 'ENV=(TDPO_OPTFILE=/oracle/tsm/client/oracle/bin/tdpo.opt)';
5> set newname for datafile 1 to '/oracle/app/oracle/oradata/ora9i/system.dbf';
6> set newname for datafile 2 to '/oracle/app/oracle/oradata/ora9i/undo1.dbf';
7> set newname for datafile 3 to '/oracle/app/oracle/oradata/ora9i/cwmlite.dbf';
8> set newname for datafile 4 to '/oracle/app/oracle/oradata/ora9i/drsys.dbf';
9> set newname for datafile 5 to '/oracle/app/oracle/oradata/ora9i/data4_1_01.dbf';
10> set newname for datafile 6 to '/oracle/app/oracle/oradata/ora9i/tools.dbf';
11> set newname for datafile 7 to '/oracle/app/oracle/oradata/ora9i/users.dbf';
12> set newname for datafile 8 to '/oracle/app/oracle/oradata/ora9i/xdb.dbf';
13> set newname for datafile 9 to '/oracle/app/oracle/oradata/ora9i/data10_1_01.dbf';
14> set newname for datafile 10 to '/oracle/app/oracle/oradata/ora9i/data10_1_02.dbf';
15> set newname for datafile 11 to '/oracle/app/oracle/oradata/ora9i/spaces.dbf';
16> set newname for datafile 12 to '/oracle/app/oracle/oradata/ora9i/data1_1_01.dbf';
17> restore database;
18> switch datafile all;
19> release channel t1;
20> }
executing command: SET ARCHIVELOG DESTINATION
executing command: SET until clause
ANS0101E NLInit: Unable to open message repository '/usr/tivoli/tsm/client/oracle/bin64/zh_CN/tdpo.cat'.
allocated channel: t1
channel t1: sid=14 devtype=SBT_TAPE
channel t1: Tivoli Data Protection for Oracle: version 5.2.0.0
executing command: SET NEWNAME
executing command: SET NEWNAME
executing command: SET NEWNAME
executing command: SET NEWNAME
executing command: SET NEWNAME
executing command: SET NEWNAME
executing command: SET NEWNAME
executing command: SET NEWNAME
executing command: SET NEWNAME
executing command: SET NEWNAME
executing command: SET NEWNAME
executing command: SET NEWNAME
Starting restore at 20-JUN-06
channel t1: starting datafile backupset restore
channel t1: specifying datafile(s) to restore from backup set
restoring datafile 00001 to /oracle/app/oracle/oradata/ora9i/system.dbf
restoring datafile 00002 to /oracle/app/oracle/oradata/ora9i/undo1.dbf
restoring datafile 00004 to /oracle/app/oracle/oradata/ora9i/drsys.dbf
restoring datafile 00007 to /oracle/app/oracle/oradata/ora9i/users.dbf
restoring datafile 00010 to /oracle/app/oracle/oradata/ora9i/data10_1_02.dbf
restoring datafile 00012 to /oracle/app/oracle/oradata/ora9i/data1_1_01.dbf
channel t1: restored backup piece 1
piece handle=bk_ora_3152_1_591497170 tag=DB_LEVEL0 params=NULL
channel t1: restored backup piece 2
piece handle=bk_ora_3152_2_591497170 tag=DB_LEVEL0 params=NULL
channel t1: restore complete
channel t1: starting datafile backupset restore
channel t1: specifying datafile(s) to restore from backup set
restoring datafile 00003 to /oracle/app/oracle/oradata/ora9i/cwmlite.dbf
restoring datafile 00005 to /oracle/app/oracle/oradata/ora9i/data4_1_01.dbf
restoring datafile 00006 to /oracle/app/oracle/oradata/ora9i/tools.dbf
restoring datafile 00008 to /oracle/app/oracle/oradata/ora9i/xdb.dbf
restoring datafile 00009 to /oracle/app/oracle/oradata/ora9i/data10_1_01.dbf
restoring datafile 00011 to /oracle/app/oracle/oradata/ora9i/spaces.dbf
channel t1: restored backup piece 1
piece handle=bk_ora_3153_1_591497170 tag=DB_LEVEL0 params=NULL
channel t1: restored backup piece 2
piece handle=bk_ora_3153_2_591497170 tag=DB_LEVEL0 params=NULL
channel t1: restore complete
Finished restore at 20-JUN-06
datafile 1 switched to datafile copy
input datafilecopy recid=13 stamp=593610382 filename=/oracle/app/oracle/oradata/ora9i/system.dbf
datafile 2 switched to datafile copy
input datafilecopy recid=14 stamp=593610382 filename=/oracle/app/oracle/oradata/ora9i/undo1.dbf
datafile 3 switched to datafile copy
input datafilecopy recid=15 stamp=593610382 filename=/oracle/app/oracle/oradata/ora9i/cwmlite.dbf
datafile 4 switched to datafile copy
input datafilecopy recid=16 stamp=593610382 filename=/oracle/app/oracle/oradata/ora9i/drsys.dbf
datafile 5 switched to datafile copy
input datafilecopy recid=17 stamp=593610382 filename=/oracle/app/oracle/oradata/ora9i/data4_1_01.dbf
datafile 6 switched to datafile copy
input datafilecopy recid=18 stamp=593610382 filename=/oracle/app/oracle/oradata/ora9i/tools.dbf
datafile 7 switched to datafile copy
input datafilecopy recid=19 stamp=593610382 filename=/oracle/app/oracle/oradata/ora9i/users.dbf
datafile 8 switched to datafile copy
input datafilecopy recid=20 stamp=593610382 filename=/oracle/app/oracle/oradata/ora9i/xdb.dbf
datafile 9 switched to datafile copy
input datafilecopy recid=21 stamp=593610382 filename=/oracle/app/oracle/oradata/ora9i/data10_1_01.dbf
datafile 10 switched to datafile copy
input datafilecopy recid=22 stamp=593610382 filename=/oracle/app/oracle/oradata/ora9i/data10_1_02.dbf
datafile 11 switched to datafile copy
input datafilecopy recid=23 stamp=593610382 filename=/oracle/app/oracle/oradata/ora9i/spaces.dbf
datafile 12 switched to datafile copy
input datafilecopy recid=24 stamp=593610382 filename=/oracle/app/oracle/oradata/ora9i/data1_1_01.dbf
RMAN>
本来archivelog文件也一样可以使用until time来找的，但是不知道为什么，我采用这样的语句始终报错，只好转换成logseq来做了：
RMAN> run {
2> set archivelog destination to '/test/arch/';
3> set until time "to_date('06/13/06 13:00:00','MM/DD/YY HH24:MI:SS')" ;
4> allocate channel t1 type 'sbt_tape' parms 'ENV=(TDPO_OPTFILE=/oracle/tsm/client/oracle/bin/tdpo.opt)';
5> restore archivelog until time "to_date('06/13/06 13:00:00','MM/DD/YY HH24:MI:SS')" ;
6> release channel t1;
7> }
executing command: SET ARCHIVELOG DESTINATION
executing command: SET until clause
ANS0101E NLInit: Unable to open message repository '/usr/tivoli/tsm/client/oracle/bin64/zh_CN/tdpo.cat'.
allocated channel: t1
channel t1: sid=11 devtype=SBT_TAPE
channel t1: Tivoli Data Protection for Oracle: version 5.2.0.0
Starting restore at 20-JUN-06
archive log thread 1 sequence 1 is already on disk as file /test/arch/arch0000000001.arc
archive log thread 1 sequence 2 is already on disk as file /test/arch/arch0000000002.arc
archive log thread 1 sequence 3 is already on disk as file /test/arch/arch0000000003.arc
archive log thread 1 sequence 4 is already on disk as file /test/arch/arch0000000004.arc
archive log thread 1 sequence 5 is already on disk as file /test/arch/arch0000000005.arc
archive log thread 1 sequence 6 is already on disk as file /test/arch/arch0000000006.arc
当然，这里没有使用from time限定，但是后来我使用了from time还是一样报错。
使用RMAN查询对应日期的logseq号：
RMAN> list backup of archivelog all;
List of Backup Sets
===================
BS Key Size Device Type Elapsed Time Completion Time
------- ---------- ----------- ------------ -------------------
41 643K SBT_TAPE 00:02:58 2005-09-26:18:55:50
BP Key: 42 Status: AVAILABLE Tag: TAG20050926T185252
Piece Name: al_ora_t570048772_s2_p1
List of Archived Logs in backup set 41
Thrd Seq Low SCN Low Time Next SCN Next Time
---- ------- ---------- ------------------- ---------- ---------
1 1 3224721 2005-09-26:18:18:24 3225856 2005-09-26:18:37:36
1 2 3225856 2005-09-26:18:37:36 3226284 2005-09-26:18:52:52
………（省略若干）…..
BS Key Size Device Type Elapsed Time Completion Time
------- ---------- ----------- ------------ -------------------
22094 1M SBT_TAPE 00:03:07 2006-05-26:21:41:16
BP Key: 22095 Status: AVAILABLE Tag: TAG20060526T213809
Piece Name: al_ora_t591485889_s3148_p1
List of Archived Logs in backup set 22094
Thrd Seq Low SCN Low Time Next SCN Next Time
---- ------- ---------- ------------------- ---------- ---------
1 1953 117583452 2006-05-26:18:38:09 117590956 2006-05-26:21:38:09
BS Key Size Device Type Elapsed Time Completion Time
------- ---------- ----------- ------------ -------------------
22107 1M SBT_TAPE 00:03:11 2006-05-27:00:41:20
BP Key: 22108 Status: AVAILABLE Tag: TAG20060527T003809
Piece Name: al_ora_t591496689_s3150_p1
List of Archived Logs in backup set 22107
Thrd Seq Low SCN Low Time Next SCN Next Time
---- ------- ---------- ------------------- ---------- ---------
1 1954 117590956 2006-05-26:21:38:09 117598466 2006-05-27:00:38:09
………（省略若干）…..
BS Key Size Device Type Elapsed Time Completion Time
------- ---------- ----------- ------------ -------------------
22864 1M SBT_TAPE 00:02:55 2006-06-13:21:40:21
BP Key: 22865 Status: AVAILABLE Tag: TAG20060613T213725
Piece Name: al_ora_t593041046_s3263_p1
List of Archived Logs in backup set 22864
Thrd Seq Low SCN Low Time Next SCN Next Time
---- ------- ---------- ------------------- ---------- ---------
1 2014 119214265 2006-06-13:18:37:26 119221819 2006-06-13:21:37:25
………（省略若干）…..
由上面可见，我们需要的archivelog，从5月27～6月13晚上，logseq号是1954~2014,SCN号是117590956～119214265，奇怪为什么是从5月27～6月13晚上？6月13晚上是用户告诉我删数据前的一个时间点，而5月27则是查出来离6月13最近的一个全备份（注意oracle恢复的时候不能往后回滚，只能向前提交archivelog来恢复）：
RMAN> list backup of database;
List of Backup Sets
===================
………（省略若干）…..
BS Key Type LV Size Device Type Elapsed Time Completion Time
------- ---- -- ---------- ----------- ------------ -------------------
22117 Incr 0 12G SBT_TAPE 00:13:27 2006-05-27:00:59:37
List of Datafiles in backup set 22117
File LV Type Ckp SCN Ckp Time Name
---- -- ---- ---------- ------------------- ----
1 0 Incr 117598811 2006-05-27:00:46:10 /dev/rsystem_rw
2 0 Incr 117598811 2006-05-27:00:46:10 /dev/rundo1_rw
4 0 Incr 117598811 2006-05-27:00:46:10 /dev/rdrsys_rw
7 0 Incr 117598811 2006-05-27:00:46:10 /dev/rusers_rw
10 0 Incr 117598811 2006-05-27:00:46:10 /dev/rdata10_1_02_rw
12 0 Incr 117598811 2006-05-27:00:46:10 /dev/rdata1_1_01_rw
Backup Set Copy #1 of backup set 22117
Device Type Elapsed Time Completion Time Tag
----------- ------------ ------------------- ---
SBT_TAPE 00:13:27 2006-05-27:00:59:37 DB_LEVEL0
List of Backup Pieces for backup set 22117 Copy #1
BP Key Pc# Status Piece Name
------- --- ----------- ----------
22119 1 AVAILABLE bk_ora_3152_1_591497170
22120 2 AVAILABLE bk_ora_3152_2_591497170
BS Key Type LV Size Device Type Elapsed Time Completion Time
------- ---- -- ---------- ----------- ------------ -------------------
22118 Incr 0 10G SBT_TAPE 00:13:43 2006-05-27:00:59:53
List of Datafiles in backup set 22118
File LV Type Ckp SCN Ckp Time Name
---- -- ---- ---------- ------------------- ----
3 0 Incr 117598812 2006-05-27:00:46:10 /dev/rcwmlite_rw
5 0 Incr 117598812 2006-05-27:00:46:10 /dev/rdata4_1_01_rw
6 0 Incr 117598812 2006-05-27:00:46:10 /dev/rtools_rw
8 0 Incr 117598812 2006-05-27:00:46:10 /dev/rxdb_rw
9 0 Incr 117598812 2006-05-27:00:46:10 /dev/rdata10_1_01_rw
11 0 Incr 117598812 2006-05-27:00:46:10 /oracle/app/oracle/product/9.2.0.6/dbs/spaces.dbf
Backup Set Copy #1 of backup set 22118
Device Type Elapsed Time Completion Time Tag
----------- ------------ ------------------- ---
SBT_TAPE 00:13:43 2006-05-27:00:59:53 DB_LEVEL0
List of Backup Pieces for backup set 22118 Copy #1
BP Key Pc# Status Piece Name
------- --- ----------- ----------
22121 1 AVAILABLE bk_ora_3153_1_591497170
22122 2 AVAILABLE bk_ora_3153_2_591497170
BS Key Type LV Size Device Type Elapsed Time Completion Time
------- ---- -- ---------- ----------- ------------ -------------------
23199 Incr 0 12G SBT_TAPE 00:11:26 2006-06-17:00:57:02
List of Datafiles in backup set 23199
File LV Type Ckp SCN Ckp Time Name
---- -- ---- ---------- ------------------- ----
1 0 Incr 119504058 2006-06-17:00:45:36 /dev/rsystem_rw
2 0 Incr 119504058 2006-06-17:00:45:36 /dev/rundo1_rw
4 0 Incr 119504058 2006-06-17:00:45:36 /dev/rdrsys_rw
7 0 Incr 119504058 2006-06-17:00:45:36 /dev/rusers_rw
10 0 Incr 119504058 2006-06-17:00:45:36 /dev/rdata10_1_02_rw
12 0 Incr 119504058 2006-06-17:00:45:36 /dev/rdata1_1_01_rw
Backup Set Copy #1 of backup set 23199
Device Type Elapsed Time Completion Time Tag
----------- ------------ ------------------- ---
SBT_TAPE 00:11:26 2006-06-17:00:57:02 DB_LEVEL0
………（省略若干）…..
有了这个logseq，就好办事了：
p570b&oracle> rman target / catalog rman/rman@rman
Recovery Manager: Release 9.2.0.6.0 - 64bit Production
Copyright (c) 1995, 2002, Oracle Corporation. All rights reserved.
connected to target database: ORACLE9I (DBID=3231295384)
connected to recovery catalog database
RMAN> run {
2> set archivelog destination to '/test/arch/';
3> allocate channel t1 type 'sbt_tape' parms 'ENV=(TDPO_OPTFILE=/oracle/tsm/client/oracle/bin/tdpo.opt)';
4> restore archivelog from logseq=1966 until logseq=2015;
5> release channel t1;
6> }
executing command: SET ARCHIVELOG DESTINATION
ANS0101E NLInit: Unable to open message repository '/usr/tivoli/tsm/client/oracle/bin64/zh_CN/tdpo.cat'.
allocated channel: t1
channel t1: sid=13 devtype=SBT_TAPE
channel t1: Tivoli Data Protection for Oracle: version 5.2.0.0
Starting restore at 2006-06-20:13:30:36
channel t1: starting archive log restore to user-specified destination
archive log destination=/test/arch/
channel t1: restoring archive log
archive log thread=1 sequence=1966
channel t1: restored backup piece 1
piece handle=al_ora_t591615479_s3175_p1 tag=TAG20060528T093759 params=NULL
channel t1: restore complete
channel t1: starting archive log restore to user-specified destination
archive log destination=/test/arch/
channel t1: restoring archive log
archive log thread=1 sequence=1967
channel t1: restored backup piece 1
piece handle=al_ora_t591626280_s3177_p1 tag=TAG20060528T123759 params=NULL
channel t1: restore complete
channel t1: starting archive log restore to user-specified destination
archive log destination=/test/arch/
channel t1: restoring archive log
archive log thread=1 sequence=1968
于是敲完这些命令就去吃中午饭了，心想吃完回来应该restore完了吧，也就60个logseq！
没想到过了2个小时才restore了10几个logseq，观察发现，每恢复一个logseq，3582都要dismount以下volume，然后再mount，location……但是这些logseq都是在同一个volume上的！按照常理，应该是一次restore完一个volume上的所有log才会dismount一个volume的，猜想是不是TSM的node的keep mount point属性影响？该属性是让一个node备份完后并不马上dismount当前volume，而是等待一段时间，当一下个备份node不是刚才的node的时候，才dismount该volume，于是中断恢复，修改该属性，问题依然。。。。
没办法，打电话给IBMer，他解释说要用SCN做条件来恢复才快，于是，改成SCN：
RMAN> run {
2> set archivelog destination to '/test/arch/';
3> allocate channel t1 type 'sbt_tape' parms 'ENV=(TDPO_OPTFILE=/oracle/tsm/client/oracle/bin/tdpo.opt)';
4> restore archivelog from scn=117841982 until scn=119229249;
5> release channel t1;
6> }
executing command: SET ARCHIVELOG DESTINATION
ANS0101E NLInit: Unable to open message repository '/usr/tivoli/tsm/client/oracle/bin64/zh_CN/tdpo.cat'.
allocated channel: t1
channel t1: sid=13 devtype=SBT_TAPE
channel t1: Tivoli Data Protection for Oracle: version 5.2.0.0
Starting restore at 2006-06-20:14:37:59
channel t1: starting archive log restore to user-specified destination
archive log destination=/test/arch/
channel t1: restoring archive log
archive log thread=1 sequence=1979
channel t1: restored backup piece 1
piece handle=al_ora_t591755880_s3201_p1 tag=TAG20060530T003800 params=NULL
还是一样的问题，再打电话问IBMer，他肯定的说是sta（TSM的存储代理，相当于veritas的media server）的TCP配置有问题，我也没高清楚是什么问题，半信半疑的禁用了STA，也就是禁用了SAN恢复，采用网络方式，然而，问题没有解决，更大的问题却来了（不过，后来的操作证明IBMer说的是对的）：
RMAN> run {
2> set archivelog destination to '/test/arch/';
3> allocate channel t1 type 'sbt_tape' parms 'ENV=(TDPO_OPTFILE=/oracle/tsm/client/oracle/bin/tdpo.opt)';
4> restore archivelog from scn=117864624 until scn=119229249;
5> release channel t1;
6> }
executing command: SET ARCHIVELOG DESTINATION
ANS0101E NLInit: Unable to open message repository '/usr/tivoli/tsm/client/oracle/bin64/zh_CN/tdpo.cat'.
allocated channel: t1
channel t1: sid=13 devtype=SBT_TAPE
channel t1: Tivoli Data Protection for Oracle: version 5.2.0.0
Starting restore at 2006-06-20:15:05:39
archive log thread 1 sequence 1982 is already on disk as file /test/arch/arch0000001982.arc
channel t1: starting archive log restore to user-specified destination
archive log destination=/test/arch/
channel t1: restoring archive log
archive log thread=1 sequence=1983
released channel: t1
RMAN-00571: ===========================================================
RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
RMAN-00571: ===========================================================
RMAN-03002: failure of restore command at 06/20/2006 15:06:07
ORA-19501: read error on file "al_ora_t591799080_s3209_p1", blockno 1 (blocksize=512)
ORA-27190: skgfrd: sbtread2 returned error
ORA-19511: Error received from media manager layer, error text:
ANS1314E (RC14) File data currently unavailable on server
RMAN> quit
居然一个log也不能恢复了！！！但是刚刚装的windows客户端备份恢复却是好好的！恢复sta的配置，不行，直接到p570a上restore archivelog（注：虽然不能在p570a上recover，但是restore是可以的，但是操作要千万小心，不然可能破坏生产环境了），还是不行，甚至再restore刚才能够restore的那些log都不行了！！！但是，在p570a上备份归档却可以，恢复刚备份的log也可以！
于是根据TSM的报错,在TSM上查原因：
tsm￡oTSMSVR>q vol A00005L2 f=d
Volume Name: A00005L2
Storage Pool Name: ORACLE
Device Class Name: LTOCLASS1
Estimated Capacity: 190,734.0
Scaled Capacity Applied:
Pct Util: 15.7
Volume Status: Filling
Access: Unavailable
Pct. Reclaimable Space: 0.0
Scratch Volume?: No
In Error State?: No
Number of Writable Sides: 1
Number of Times Mounted: 335
Write Pass Number: 1
Approx. Date Last Written: 06/20/2006 09:54:06
Approx. Date Last Read: 06/20/2006 15:14:04
Date Became Pending:
Number of Write Errors: 0
Number of Read Errors: 0
Volume Location:
Volume is MVS Lanfree Capable : No
Last Update by (administrator): ADMIN
Last Update Date/Time: 09/06/2005 11:08:08
Begin Reclaim Period:
End Reclaim Period:
A00005L2是就是恢复需要的volume，居然变成Unavailable了！难道是刚才疯狂mount把它弄坏了。。。
尝试手工把状态改回来：
tsm￡oTSMSVR>update volume A00005L2 access=READWrite
ANR2207I Volume A00005L2 updated.
tsm￡oTSMSVR>q vol A00005L2 f=d
Volume Name: A00005L2
Storage Pool Name: ORACLE
Device Class Name: LTOCLASS1
Estimated Capacity: 190,734.0
Scaled Capacity Applied:
Pct Util: 15.7
Volume Status: Filling
Access: Read/Write
Pct. Reclaimable Space: 0.0
Scratch Volume?: No
In Error State?: No
Number of Writable Sides: 1
Number of Times Mounted: 335
Write Pass Number: 1
Approx. Date Last Written: 06/20/2006 09:54:06
Approx. Date Last Read: 06/20/2006 15:14:04
Date Became Pending:
Number of Write Errors: 0
Number of Read Errors: 0
Volume Location:
Volume is MVS Lanfree Capable : No
Last Update by (administrator): ADMIN
Last Update Date/Time: 06/20/2006 16:42:32
Begin Reclaim Period:
End Reclaim Period:
居然OK了，呵呵，没搞清楚是什么原因，继续往下做吧，没想到还真的可以了，而且去掉sta的参数后，log也是一下子恢复出来了，而不是象刚才一个log要mount一次volume。
阅读(1423) | 评论(1) | 转发(0) |
