ORACLE11g R2【RAC+ASM→RAC+ASM】
ORACLE11g R2【RAC+ASM→RAC+ASM】
本演示案例所用環境：RAC+ASM+OMF
|
|
primary |
standby |
|
OS Hostname |
node1,node2 |
dgnode1,dgnode2 |
|
OS Version |
RHEL6.5 |
RHEL6.5 |
|
DB Version |
11.2.0.4 |
11.2.0.4 |
|
db_name |
stephen |
stephen |
|
db_unique_name |
stephen |
standby |
|
service_names |
stephen |
standby |
|
instance_name |
stephen1,stephen2 |
standby1,standby2 |
Primary database configure
1.啟用primary force logging
SQL> select force_logging from v$database;
SQL> alter database force logging;
2.配置redo傳輸認證
(1) tnsnames.ora
STEPHEN =
(DESCRIPTION =
(ADDRESS = (PROTOCOL = TCP)(HOST = node1-vip)(PORT = 1521))
(ADDRESS = (PROTOCOL = TCP)(HOST = node2-vip)(PORT = 1521))
(CONNECT_DATA =
(SERVER = DEDICATED)
(SERVICE_NAME = stephen)
)
)
STANDBY =
(DESCRIPTION =
(ADDRESS = (PROTOCOL = TCP)(HOST = dgnode1-vip)(PORT = 1521))
(ADDRESS = (PROTOCOL = TCP)(HOST = dgnode2-vip)(PORT = 1521))
(CONNECT_DATA =
(SERVER = DEDICATED)
(SERVICE_NAME = standby)
)
)
(2) tnsping命令解析網絡服務名,命令須返回OK.
此時，standby端未配置監聽，無法解析，可以standby配置後再測試。
$ tnsping standby
(3) 密碼文件
若無密碼文件，則在一個節點創建密碼文件並拷貝至其余節點。格式：orapw<ORACLE_SID>
cd $ORACLE_HOME/dbs
orapwd file=orapwstephen1 password=oracle ignorecase=y force=y
scp orapwstephen1 node2:$ORACLE_HOME/dbs/orapwstephen2
3.添加standby logfiles
添加standby logfile 的要求：
(1) 確保主和備數據庫上的日誌文件大小是相同的
(2) 確定備庫重做日誌文件組的適當數目
Standby logfile數 = （每個線程的logfile數＋1）* 線程數
(3) 檢查create database時指定的MAXLOGFILES和MAXLOGMEMBERS參數，可以通過trace controlfile出來查看
(4) RAC環境創建standby logfile,指定thread#創建
每個thread的standby logfile數 = 每個thread的logfile數 + 1
檢查當前環境的logfile：
set lines 200 pages 300
col member for a60
select a.thread#,a.group#,b.member,b.type,a.bytes/1024/1024 MB from v$log a,v$logfile b where a.group#=b.group#
union all
select a.thread#,a.group#,b.member,b.type,a.bytes/1024/1024 MB from v$standby_log a,v$logfile b where a.group#=b.group#;
根據查詢的logfile 信息上文添加standby logfile的要求，添加合理的standby logfile,然後用上面的sql再次查看當前環境的logfile：
alter database add standby logfile thread 1
group 5 ‘+DATA‘ size 50M,
group 6 ‘+DATA‘ size 50M,
group 7 ‘+DATA‘ size 50M;
alter database add standby logfile thread 2
group 8 ‘+DATA‘ size 50M,
group 9 ‘+DATA‘ size 50M,
group 10 ‘+DATA‘ size 50M;
4.修改primary參數文件
備份spfile：
create pfile=‘/tmp/pfile.bak‘ from spfile;
修改以下參數，其中stephen,standby分別為primary,standby的db_unqiue_name/TNS-Alias,詳見參數詳解部分：
alter system set log_archive_config=‘dg_config=(stephen,standby)‘ sid=‘*‘;
alter system set log_archive_dest_1=‘location=+DATA valid_for=(all_logfiles,all_roles) db_unique_name=stephen‘ sid=‘*‘;
alter system set log_archive_dest_2=‘service=standby lgwr async valid_for=(online_logfiles,primary_role) db_unique_name=standby‘ sid=‘*‘;
alter system set log_archive_max_processes=30 sid=‘*‘; #根據需求調整個數
--確認以下參數，若不為默認值則修改：
alter system set remote_login_passwordfile=exclusive scope=spfile sid=‘*‘;
# remote_login_passwordfile=exclusive/shared
alter system set log_archive_dest_state_1=enable sid=‘*‘;
alter system set log_archive_dest_state_2=enable sid=‘*‘;
--可修改歸檔日誌格式，與datafile區分：
alter system set log_archive_format=‘%t_%s_%r.arc‘ scope=spfile sid=‘*‘;
--備角色參數：
alter system set fal_server=standby sid=‘*‘;
alter system set standby_file_management=auto sid=‘*‘;
alter system set db_file_name_convert=‘+DATA/STANDBY‘,‘+DATA/STEPHEN‘ scope=spfile sid=‘*‘;
alter system set log_file_name_convert=‘+DATA/STANDBY‘,‘+DATA/STEPHEN‘ scope=spfile sid=‘*‘;
註：以上參數，指定spfile修改的，在實例重啟後生效。
采用OMF時，主、備庫均只能使用一個ASM磁盤組,否則禁用OMF：
alter system reset db_create_file_dest sid=‘*‘;
alter system reset db_create_online_log_dest_n sid=‘*‘; #n為數值變量
alter system reset db_recovery_file_dest sid=‘*‘;
5.打開歸檔模式
首先確認當前數據庫是否是歸檔模式，如果不是，則打開歸檔模式。
archive log list;
srvctl stop database -d stephen
srvctl start database -d stephen -o mount
alter database archivelog; --在一個節點執行即可
alter database open; --各節點執行
6.備份primary database
rman target / <<EOF
run{allocate channel c1 type disk;
allocate channel c2 type disk;
backup full database format ‘/backup/full_%U.bak‘ plus archivelog;
sql ‘alter system switch logfile‘;
backup current controlfile for standby format ‘/backup/c_%U.bak‘;
release channel c1;
release channel c2;
}
EOF
Standby database configure
前期準備
首先安裝CRS,RDBMS軟件，配置ASM，netca創建監聽。
1.配置各節點環境變量
export ORACLE_BASE=/oracle/app/oracle
export ORACLE_HOME=$ORACLE_BASE/product/11.2.0/db_1
export ORACLE_SID=standby1 #根據節點修改
export PATH=$ORACLE_HOME/bin:$PATH
export LD_LIBRARY_PATH=$ORACLE_HOME/lib #不同平臺變量名不一樣
2.配置redo傳輸認證
(1) tnsnames.ora
配置各節點的$ORACLE_HOME/network/admin/tnsnames.ora文件：
STEPHEN =
(DESCRIPTION =
(ADDRESS = (PROTOCOL = TCP)(HOST = node1-vip)(PORT = 1521))
(ADDRESS = (PROTOCOL = TCP)(HOST = node2-vip)(PORT = 1521))
(CONNECT_DATA =
(SERVER = DEDICATED)
(SERVICE_NAME = stephen)
)
)
STANDBY =
(DESCRIPTION =
(ADDRESS = (PROTOCOL = TCP)(HOST = dgnode1-vip)(PORT = 1521))
(ADDRESS = (PROTOCOL = TCP)(HOST = dgnode2-vip)(PORT = 1521))
(CONNECT_DATA =
(SERVER = DEDICATED)
(SERVICE_NAME = standby)
)
)
重載各節點監聽：
lsnrctl reload
(2) tnsping命令解析網絡服務名,命令須返回OK.
$ tnsping stephen
(3) 密碼文件
使用主庫的密碼文件：
scp $ORACLE_HOME/dbs/orapwstephen1 dgnode1:$ORACLE_HOME/dbs/orapwstandby1
scp $ORACLE_HOME/dbs/orapwstephen1 dgnode2:$ORACLE_HOME/dbs/orapwstandby2
3.創建目錄
在各節點創建相關目錄：
mkdir -p $ORACLE_BASE/admin/standby/{adump,dpdump,pfile}
創建ASM目錄：
export ORACLE_SID=+ASM1
SQL> alter diskgroup data add directory ‘+DATA/STANDBY‘;
4.參數文件
用primary的pfile加以修改，以保證某些參數與primary保持一致，註意主備庫內存的大小適當的調整內存相關參數。
scp node1:/tmp/pfile.bak $ORACLE_HOME/dbs/initstandby1.ora
修改pfile為需要的standby pfile,特別註意以下事例的參數設定：
*.audit_file_dest=‘/u01/app/oracle/admin/standby/adump‘
*.cluster_database_instances=2
*.cluster_database=true
*.db_create_file_dest=‘+DATA‘
*.db_name=‘stephen‘
*.db_unique_name=‘standby‘
*.service_names=‘standby‘
standby1.instance_name=‘standby1‘
standby2.instance_name=‘standby2‘
standby2.instance_number=2
standby1.instance_number=1
*.remote_listener=‘cluster-scan:1521‘
standby1.local_listener=‘(ADDRESS=(PROTOCOL=TCP)(HOST=dgnode1-vip)(PORT=1521))‘
standby2.local_listener=‘(ADDRESS=(PROTOCOL=TCP)(HOST=dgnode2-vip)(PORT=1521))‘
*.remote_login_passwordfile=‘exclusive‘
standby2.thread=2
standby1.thread=1
*.undo_management=‘AUTO‘
standby1.undo_tablespace=‘UNDOTBS1‘
standby2.undo_tablespace=‘UNDOTBS2‘
*.log_archive_config=‘dg_config=(stephen,standby)‘
*.log_archive_dest_1=‘location=+DATA valid_for=(all_logfiles,all_roles) db_unique_name=standby‘
*.log_archive_dest_2=‘service=stephen lgwr async valid_for=(online_logfiles,primary_role) db_unique_name=stephen‘
*.log_archive_max_processes=30 #根據需求調整個數
*.remote_login_passwordfile=exclusive # exclusive/shared
*.log_archive_dest_state_1=enable
*.log_archive_dest_state_2=enable
*.log_archive_format=‘%t_%s_%r.arc‘
*.fal_server=stephen
*.standby_file_management=auto
*.db_file_name_convert=‘+DATA/STEPHEN‘,‘+DATA/STANDBY‘
*.log_file_name_convert=‘+DATA/STEPHEN‘,‘+DATA/STANDBY‘
以pfile啟動數據庫實例到nomount階段,只需要啟動一個實例用來做恢復即可：
SQL> startup nomount pfile=‘?/dbs/initstandby1.ora‘;
註意：以下恢復操作，只在一個節點做。
5.恢復standby controlfile
SQL> startup nomount;
RMAN> restore standby controlfile to ‘+DATA‘ from ‘/backup/c_0aqn43be_1_1.bak‘;
修改control_files參數執行controlfile位置:
alter system set control_files=‘+DATA/standby/controlfile/current.257.896669437‘ scope=spfile sid=‘*‘;
重啟實例：
shutdown immediate;
startup nomount pfile=‘?/dbs/initstandby1.ora‘;
alter database mount standby database;
6.恢復數據文件
查詢primary database的數據文件信息：
set lines 300 pages 300
col name for a60
col member for a60
select file#,name from v$datafile
union all
select file#,name from v$tempfile;
啟動standby到mount階段，恢復數據文件：
RMAN> run{
set newname for datafile 1 to ‘+DATA‘;
set newname for datafile 2 to ‘+DATA‘;
set newname for datafile 3 to ‘+DATA‘;
set newname for datafile 4 to ‘+DATA‘;
set newname for datafile 5 to ‘+DATA‘;
set newname for tempfile 1 to ‘+DATA‘;
restore database;
switch datafile all;
switch tempfile all;
recover database;
}
查看standby database log file:
set lines 200 pages 300
col member for a60
select a.thread#,a.group#,b.member,b.type,a.bytes/1024/1024 MB from v$log a,v$logfile b where a.group#=b.group#
union all
select a.thread#,a.group#,b.member,b.type,a.bytes/1024/1024 MB from v$standby_log a,v$logfile b where a.group#=b.group#;
7.註冊OCR
註冊DB,Instance到OCR：
srvctl add database -d standby -o $ORACLE_HOME
srvctl add instance -d standby -i standby1 -n dgnode1
srvctl add instance -d standby -i standby2 -n dgnode2
查看資源狀態：
crsctl status res -t -w "TYPE = ora.database.type"
--------------------------------------------------------------------------------
NAMETARGET STATE SERVER STATE_DETAILS
--------------------------------------------------------------------------------
Cluster Resources
--------------------------------------------------------------------------------
ora.standby.db
1 OFFLINE OFFLINE
2 OFFLINE OFFLINE
啟用DB跟隨CRS啟動：
srvctl enable database -d standby
8.創建spfile
SQL> create spfile=‘+DATA‘ from pfile=‘?/dbs/initstandby1.ora‘;
關閉數據庫實例，修改各節點pfile,然後以spfile啟動數據庫：
SQL> shutdown immediate
通過asmcmd命令查看spfile的位置
asmcmd find +DATA/standby spfile*
+DATA/standby/parameterfile/spfile.256.896665481
修改各節點pfile：
cd $ORACLE_HOME/dbs
echo "spfile=‘+DATA/standby/parameterfile/spfile.256.896665481">initstandby1.ora
echo "spfile=‘+DATA/standby/parameterfile/spfile.256.896665481">initstandby2.ora
以spfile啟動實例到mount階段(Data Guard）或open read only階段(Active Data Guard)：
--啟動到mount：
srvctl start database -d standby -o mount
--啟動到open read only：
srvctl start database -d standby -o ‘read only‘
查看資源狀態：
crsctl status res -t -w "TYPE = ora.database.type"
--------------------------------------------------------------------------------
NAMETARGET STATE SERVER STATE_DETAILS
--------------------------------------------------------------------------------
Cluster Resources
--------------------------------------------------------------------------------
ora.standby.db
1 OFFLINE OFFLINE dgnode 1 Open,Readonly
2 OFFLINE OFFLINE dgnode 2 Open,Readonly
如果狀態顯示為Open,此為ORACLE BUG,參考MOS DOC 1622876.1
9.應用日誌
應用日誌只需要在一個節點執行即可,日誌應用可選擇應用archive log或redo log.
應用archive log：
recover managed standby database disconnect;
實時應用redo logfile：
recover managed standby database using current logfile disconnect;
取消應用日誌：
recover managed standby database cancel;
打開數據庫至read only模式並開始實時復制：
recover managed standby database using current logfile disconnect;
10.觀察日誌傳輸
跟蹤primary,standby database的alert log,觀察是否有錯誤發生，也可以看到應用日誌的相關信息。
確認standby是否應用日誌：
--primary端多切幾次日誌，觀察alert log信息。
alter system switch logfile;
alter system switch logfile;
--觀察主備庫日誌序列號
archive log list;
--primary端查詢v$archived_log視圖，確認日誌是否被應用：
set lines 300 pages 300
col name for a20
select name,dest_id,thread#,sequence#,standby_dest,applied,registrar,completion_time from v$archived_log
where standby_dest=‘YES‘
order by thread#,sequence#;
--primary端查詢primary,standby的最大日誌序列號是否一致：
select ‘Primary :‘ "DB Role",thread#,max(sequence#)
from v$archived_log
where standby_dest=‘NO‘
group by thread#
union
select ‘Standby :‘ "DB Role",thread#,max(sequence#)
from v$archived_log
where standby_dest=‘YES‘ and applied=‘YES‘
group by thread#
order by thread#;
ORACLE11g R2【RAC+ASM→RAC+ASM】