数据库从单实例迁移到Oracle RAC的详细步骤
本文详细介绍了将数据库从单实例迁移到Oracle RAC的全过程，包括准备阶段、备份数据库、检查ASM磁盘组空间、创建RAC实例、恢复数据库、数据迁移、参数调整、初始化实例等关键步骤。
Migrate database from single instance to Oracle RAC
# Preparation before restore
# Backup database
# check listener
# check ASM disk group free space is sufficient.
# the following steps was completed sucessful on Oracle 10g RAC + Suse Linux 10.
Synopsis:
source DB : GOBO1 on file system
Target DB : GOBO1 on RAC +ASM
Target Instance: GOBO1A, GOBO1B
Target node: bo2dbp, bo2dbs
Source Env: Oracle 10g(10.2.0.3) + Suse 10
Target Env: Oracle 10g(10.2.0.3) RAC + ASM + RAW + Suse 10
ORA_CRS_HOME=/u01/oracle/crs
ORA_ASM_HOME=/u01/oracle/asm
ORACLE_BASE=/u01/oracle
ORACLE_HOME=/u01/oracle/db
Step 1
# restore spfile
export ORACLE_SID=GOBO1A
rman target /
startup nomount;
restore spfile to pfile '/u01/oracle/db/dbs/initGOBO1A.ora'
from '<dir>';
shutdown immediate;
Step 2
# create directory for instance on local file system.(two nodes)
export ORACLE_SID=GOBO1
mkdir -p /u01/oracle/admin/${ORACLE_SID}/{bdump,cdump,udump,adump}
Step 3
#Modify pfile
#Remove original path and or change them to new path on target server.
#Add new item for cluster
cp initGOBO1A.ora initGOBO1A.ora.bak
vi initGOBO1A.ora
GOBO1.__db_cache_size=230686720
GOBO1.__java_pool_size=4194304
GOBO1.__large_pool_size=4194304
GOBO1.__shared_pool_size=289406976
GOBO1.__streams_pool_size=0
*.compatible='10.2.0.3.0'
#*.control_files='/u02/database/GOBO1/controlf/cntl1GOBO1.ctl','/u02/database/GOBO1/controlf/cntl2GOBO1.ctl',
#'/u02/database/GOBO1/controlf/cntl3GOBO1.ctl'
#*.core_dump_dest='/u02/database/GOBO1/cdump'
*.db_block_size=8192
*.db_domain=''
*.db_file_multiblock_read_count=16
*.DB_FLASHBACK_RETENTION_TARGET=2880
*.db_name='GOBO1'
#*.db_recovery_file_dest='/u02/database/GOBO1/flash_recovery_area'
*.db_recovery_file_dest_size=4G
*.dispatchers='(PROTOCOL=TCP) (SERVICE=GOBO1XDB)'
*.job_queue_processes=10
#*.LOG_ARCHIVE_DEST_1='LOCATION=/u02/database/GOBO1/archive/'
*.log_archive_format='arch_%r_%t_%s.arc'
*.open_cursors=300
*.pga_aggregate_target=199229440
*.processes=1000
*.remote_login_passwordfile='EXCLUSIVE'
*.sga_target=512M
*.shared_pool_size=256M
#*.undo_management='AUTO'
#*.undo_tablespace='UNDOTBS1'
#*.user_dump_dest='/u02/database/GOBO1/udump'
#*.UTL_FILE_DIR='/u02/database/GOBO1/udump'
#*.background_dump_dest='/u02/database/GOBO1/bdump'
#Added new dump directory
*.core_dump_dest='/u01/oracle/admin/GOBO1/cdump'
*.user_dump_dest='/u01/oracle/admin/GOBO1/udump'
*.UTL_FILE_DIR='/u01/oracle/admin/GOBO1/udump'
*.background_dump_dest='/u01/oracle/admin/GOBO1/bdump'
*.audit_file_dest='/u01/oracle/admin/GOBO1/adump'
*.cluster_database = TRUE
*.cluster_database_instances = 2
*.undo_management='AUTO'
*.control_files='+DG1/GOBO1/controlf/cntl1GOBO1.ctl','+DG1/GOBO1/controlf/cntl2GOBO1.ctl'
GOBO1A.undo_tablespace='UNDOTBS1'
GOBO1A.instance_name=GOBO1A
GOBO1A.instance_number=1
GOBO1A.thread=1
#GOBO1A.local_listener=<LISTENERNAME>_<HOSTNAME1>
GOBO1B.undo_tablespace='UNDOTBS2'
GOBO1B.instance_name=GOBO1B
GOBO1B.instance_number=2
#GOBO1B.thread=2
#GOBO1B.local_listener=<LISTENERNAME>_<HOSTNAME2>
*.db_create_file_dest='+DG2'
*.db_recovery_file_dest='+REV'
*.log_archive_dest_1='LOCATION=+REV/GOBO1/archivelog'
Step 4
#Create directory on ASM for new database
#Currently, seperate different file type to different disk group.
export ORACLE_SID=+ASM1
asmcmd
cd +DG1
mkdir GOBO1
cd GOBO1
mkdir controlf parameterf onlinelog
cd +DG2
mkdir GOBO1
cd GOBO1
mkdir datafile
Step 5
#Create password on 2 nodes.
$ORACLE_HOME/dbs/orapwd file=?/dbs/orapwGOBO1A password=oracle entries=10
$ORACLE_HOME/dbs/orapwd file=?/dbs/orapwGOBO1B password=oracle entries=10
Step 6
#Generate spfile from pfile
export ORACLE_SID=GOBO1A
sqlplus / as sysdba
startup nomount;
create spfile='+DG1/GOBO1/parameterf/spfileGOBO1.ora' from pfile='/u01/oracle/db/dbs/initGOBO1A.ora'
shutdown immediate;
echo "SPFILE='+DG1/GOBO1/parameterf/spfileGOBO1.ora'" >/u01/oracle/db/dbs/initGOBO1A.ora # on node A ( two nodes)
echo "SPFILE='+DG1/GOBO1/parameterf/spfileGOBO1.ora'" >/u01/oracle/db/dbs/initGOBO1B.ora # on node B ( two nodes)
startup nomount;
show parameter spfile;
Step 7
#Restore controlfile from autobackup
export ORACLE_SID=GOBO1A
$ORACLE_HOME/bin/rman target /
restore controlfile from '<dir>';
sql 'alter database mount';
Step 8
#Check datafile path and convert to ASM
#Restore Database
sys@GOBO1> select file_id,file_name from dba_data_files order by 1;
FILE_ID FILE_NAME
---------- ------------------------------------------------------------
1 /u02/database/GOBO1/oradata/sysGOBO1.dbf
2 /u02/database/GOBO1/undo/undotbsGOBO1.dbf
3 /u02/database/GOBO1/oradata/sysauxGOBO1.dbf
4 /u02/database/GOBO1/undo/undotbsGOBO12.dbf
5 /u02/database/GOBO1/oradata/GOBO1_account_tbl.dbf
6 /u02/database/GOBO1/oradata/GOBO1_stock_tbl.dbf
7 /u02/database/GOBO1/oradata/GOBO1_stock_l_tbl.dbf
8 /u02/database/GOBO1/oradata/GOBO1_tx_tbl.dbf
9 /u02/database/GOBO1/oradata/GOBO1_users_tbl.dbf
10 /u02/database/GOBO1/oradata/GOBO1_account_idx.dbf
11 /u02/database/GOBO1/oradata/GOBO1_stock_idx.dbf
12 /u02/database/GOBO1/oradata/GOBO1_stock_l_idx.dbf
13 /u02/database/GOBO1/oradata/GOBO1_tx_idx.dbf
14 /u02/database/GOBO1/oradata/GOBO1_users_idx.dbf
15 /u02/database/GOBO1/oradata/GOBO1_IES_IDX.DBF
16 /u02/database/GOBO1/oradata/GOBO1_IES_TBL.DBF
17 /u02/database/GOBO1/oradata/GOBO1_import_idx.dbf
18 /u02/database/GOBO1/oradata/GOBO1_import_tbl.dbf
19 /u02/database/GOBO1/oradata/GOBO1_fix_tx_idx.dbf
20 /u02/database/GOBO1/oradata/GOBO1_fix_tx_tbl.dbf
21 /u02/database/GOBO1/oradata/GOBO1_fix_users_idx.dbf
22 /u02/database/GOBO1/oradata/GOBO1_fix_users_tbl.dbf
23 /u02/database/GOBO1/oradata/xxxx_ipo_tbl.dbf
24 /u02/database/GOBO1/oradata/xxxx_ipo_idx.dbf
catalog start with '/install_source/rman_bak';
run {
set newname for datafile 1 to '+DG2/GOBO1/datafile/sysGOBO1.dbf';
set newname for datafile 2 to '+DG2/GOBO1/datafile/undotbsGOBO1.dbf';
set newname for datafile 3 to '+DG2/GOBO1/datafile/sysauxGOBO1.dbf';
set newname for datafile 4 to '+DG2/GOBO1/datafile/undotbsGOBO12.dbf';
set newname for datafile 5 to '+DG2/GOBO1/datafile/GOBO1_account_tbl.dbf';
set newname for datafile 6 to '+DG2/GOBO1/datafile/GOBO1_stock_tbl.dbf';
set newname for datafile 7 to '+DG2/GOBO1/datafile/GOBO1_stock_l_tbl.dbf';
set newname for datafile 8 to '+DG2/GOBO1/datafile/GOBO1_tx_tbl.dbf';
set newname for datafile 9 to '+DG2/GOBO1/datafile/GOBO1_users_tbl.dbf';
set newname for datafile 10 to '+DG2/GOBO1/datafile/GOBO1_account_idx.dbf';
set newname for datafile 11 to '+DG2/GOBO1/datafile/GOBO1_stock_idx.dbf';
set newname for datafile 12 to '+DG2/GOBO1/datafile/GOBO1_stock_l_idx.dbf';
set newname for datafile 13 to '+DG2/GOBO1/datafile/GOBO1_tx_idx.dbf';
set newname for datafile 14 to '+DG2/GOBO1/datafile/GOBO1_users_idx.dbf';
set newname for datafile 15 to '+DG2/GOBO1/datafile/GOBO1_IES_IDX.DBF';
set newname for datafile 16 to '+DG2/GOBO1/datafile/GOBO1_IES_TBL.DBF';
set newname for datafile 17 to '+DG2/GOBO1/datafile/GOBO1_import_idx.dbf';
set newname for datafile 18 to '+DG2/GOBO1/datafile/GOBO1_import_tbl.dbf';
set newname for datafile 19 to '+DG2/GOBO1/datafile/GOBO1_fix_tx_idx.dbf';
set newname for datafile 20 to '+DG2/GOBO1/datafile/GOBO1_fix_tx_tbl.dbf';
set newname for datafile 21 to '+DG2/GOBO1/datafile/GOBO1_fix_users_idx.dbf';
set newname for datafile 22 to '+DG2/GOBO1/datafile/GOBO1_fix_users_tbl.dbf';
set newname for datafile 23 to '+DG2/GOBO1/datafile/xxxx_ipo_tbl.dbf';
set newname for datafile 24 to '+DG2/GOBO1/datafile/xxxx_ipo_idx.dbf';
restore database;
switch datafile all;
switch tempfile all;
}
Step 9
#Recover database
recover database;
Step 10
#Handle online redo log
sys@GOBO1> select * from v$logfile;
GROUP# STATUS TYPE MEMBER IS_
---------- ------- ------- ------------------------------------------------------------ ---
1 ONLINE /u02/database/GOBO1/redolog/log1aGOBO1.log NO
1 ONLINE /u02/database/GOBO1/redolog/log1bGOBO1.log NO
3 ONLINE /u02/database/GOBO1/redolog/log3aGOBO1.log NO
3 ONLINE /u02/database/GOBO1/redolog/log3bGOBO1.log NO
2 ONLINE /u02/database/GOBO1/redolog/log2aGOBO1.log NO
2 ONLINE /u02/database/GOBO1/redolog/log2bGOBO1.log NO
alter database rename file '/u02/database/GOBO1/redolog/log1aGOBO1.log' to '+DG1/GOBO1/onlinelog/log1aGOBO1.log';
alter database rename file '/u02/database/GOBO1/redolog/log1bGOBO1.log' to '+DG1/GOBO1/onlinelog/log1bGOBO1.log';
alter database rename file '/u02/database/GOBO1/redolog/log3aGOBO1.log' to '+DG1/GOBO1/onlinelog/log3aGOBO1.log';
alter database rename file '/u02/database/GOBO1/redolog/log3bGOBO1.log' to '+DG1/GOBO1/onlinelog/log3bGOBO1.log';
alter database rename file '/u02/database/GOBO1/redolog/log2aGOBO1.log' to '+DG1/GOBO1/onlinelog/log2aGOBO1.log';
alter database rename file '/u02/database/GOBO1/redolog/log2bGOBO1.log' to '+DG1/GOBO1/onlinelog/log2bGOBO1.log';
#Add online log for instance 2
alter database add logfile thread 2 group 4
('+DG1/GOBO1/onlinelog/log4aGOBO1.log','+DG1/GOBO1/onlinelog/log4bGOBO1.log') size 20M;
alter database add logfile thread 2 group 5
('+DG1/GOBO1/onlinelog/log5aGOBO1.log','+DG1/GOBO1/onlinelog/log5bGOBO1.log') size 20M;
alter database add logfile thread 2 group 6
('+DG1/GOBO1/onlinelog/log6aGOBO1.log','+DG1/GOBO1/onlinelog/log6bGOBO1.log') size 20M;
Step 11
#open the database with resetlogs
alter database open resetlogs; --> if failed, recover database using backup controlfile until cancel by sqlplus;
Step 12
#modify parameter
alter system set thread=1 scope=spfile sid='GOBO1A';
alter system set thread=2 scope=spfile sid='GOBO1B';
alter database enable thread 2;
Step 13
#add undo tablspace for instance 2
alter system set undo_tablespace='UNDOTBS2' scope=spfile sid='GOBO1B';
-->if current db has no undotbs2,create it firstly as follows
create undo tablespace UNDOTBS2 datafile '+DG2/GOBO1/datafile/undotbs02.dbf' size 500m autoextnd on;
Step 14
# add temporary tablespace and datafile
col file_name format a55
select file_name,tablespace_name,bytes/1022/1024 from dba_temp_files;
alter tablespace temp add tempfile '+DG2/GOBO1/datafile/temp.dbf' size 50m;
alter tablespace goex_temp add tempfile '+DG2/GOBO1/datafile/goex_temp.dbf' size 50m;
Step 15
# check default temporary tablespace for all users. If that are different from orginal database, amend them.
select username,default_tablespace,temporary_tablespace from dba_users;
Step 16
#check parameter
select * from v$option where parameter = 'Real Application Clusters';
show parameter cluster;
show parameter thread;
show parameter instance_number;
Step 17
# restart instance 1
# check alert log file
Step 18
# start instance 2
# check alert log file
Step 19
# check all instance is fine.
select instance_number,instance_name,host_name from gv$instance;
Step 20
#create cluster database specific views within the existing instance
$ORACLE_HOME/rdbms/admin/catclust.sql
Step 21
# configure listener
by netca
ps -ef | grep lsnr
crs_stat -t #check listener
Step 22
#Add configuration to crs
srvctl add database -d GOBO1 -o $ORACLE_HOME -p +DG1/GOBO1/parameterf/spfileGOBO1.ora
srvctl add instance -d GOBO1 -i GOBO1A -n bo2dbp
srvctl add instance -d GOBO1 -i GOBO1B -n bo2dbs
srvctl modify instance -d GOBO1 -i GOBO1A -s +ASM1
srvctl modify instance -d GOBO1 -i GOBO1B -s +ASM2
crs_stat -t
Step 23
#Restart database
#Author : Robinson
#Blog : http://blog.csdn.net/robinson_0612
srvctl start database -d GOBO1
更多参考：
有关Oracle 网络配置相关基础以及概念性的问题请参考：
配置ORACLE 客户端连接到数据库
配置非默认端口的动态服务注册
配置sqlnet.ora限制IP访问Oracle
Oracle 监听器日志配置与管理
设置 Oracle 监听器密码(LISTENER)
Oracle RAC 监听配置
ORACLE RAC 下非缺省端口监听配置(listener.ora tnsnames.ora)
Oracle RAC 客户端连接负载均衡(Load Balance)
有关基于用户管理的备份和备份恢复的概念请参考
Oracle 基于备份控制文件的恢复(unsing backup controlfile)
有关RMAN的备份恢复与管理请参考
有关ORACLE体系结构请参考
扫一扫
2499
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
