CentOS 5.11下Oracle 11G R2 Dataguard搭建

Datagard算是Oracle企業版的一種容災方案，在企業中廣泛應用，我就將搭建過程記錄下來以作備用。

主機名數據庫版本實例名 IP

db1 Oracle 11G R2 member 172.16.1.250

db2 Oracle 11G R2 member 172.16.1.251

默認情況下以上都已經安裝好了Oracle數據庫，但是隻在db1上建立了數據庫和監聽，db2只安裝Oracle軟件不建庫，不建監聽。

打開強制歸檔日誌
增加standby日誌組
修改主備啓動參數
密碼文件的處理
修改監聽
複製監聽文件、參數文件、密碼文件到備庫
創建備庫控制文件
複製主庫數據文件和日誌文件到備庫
初始化及配置備庫做standby
DataGuard測試
主備切換測試

基礎工作:

a.安裝CentOS 5.11 x86_64,關閉selinux，iptables，自動對時

b.安裝Oracle 11G R2,db1安裝軟件、監聽及建庫,db2只安裝軟件不建庫

可以參考:http://fengwan.blog.51cto.com/508652/1330122

在db1的/etc/hosts裏增加

127.0.0.1 db1

172.16.1.251 db2

在db2的/etc/hosts裏增加

127.0.0.1 db2

172.16.1.250 db1

打開強制歸檔(db1)

(db1)SQL >shutdown immediate;

(db1)SQL >startup mount;

(db1)SQL >alter database force logging;

(db1)SQL >alter database archivelog;

2.創建重做日誌組(必須要比原來的redo log多一組或多組,standby redo log是使用Real Time Apply的必要條件)

(db1)SQL> select group#,member from v$logfile;

GROUP# MEMBER

------------- -----------------------------------------------------------------

3 /opt/oracle/oradata/member/redo03.log

2 /opt/oracle/oradata/member/redo02.log

1 /opt/oracle/oradata/member/redo01.log

從上面可以看出現在已經有3組redo log.

(db1)SQL >alter database add standby logfile ('/opt/oracle/oradata/member/standby04.log') size 50m;

(db1)SQL >alter database add standby logfile ('/opt/oracle/oradata/member/standby05.log') size 50m;

(db1)SQL >alter database add standby logfile ('/opt/oracle/oradata/member/standby06.log') size 50m;

(db1)SQL >alter database add standby logfile ('/opt/oracle/oradata/member/standby07.log') size 50m;

再查下日誌組是否創建成功

(db1)SQL> select group#,member from v$logfile;

GROUP# MEMBER

------------- -----------------------------------------------------------------

3 /opt/oracle/oradata/member/redo03.log

2 /opt/oracle/oradata/member/redo02.log

1 /opt/oracle/oradata/member/redo01.log

4 /opt/oracle/oradata/member/standby04.log

5 /opt/oracle/oradata/member/standby05.log

6 /opt/oracle/oradata/member/standby06.log

7 /opt/oracle/oradata/member/standby07.log

可以看到我們創建的4組日誌

3.修改主備庫的啓動參數

生成參數文件

(db1)SQL> create pfile='/tmp/member.pfile' from spfile;

退出sqlplus,用編輯器打開/tmp/member.pfile

[oracle@db1 ~]$ vi /tmp/member.pfile

member.__db_cache_size=331350016

member.__java_pool_size=4194304

member.__large_pool_size=4194304

member.__oracle_base='/opt/oracle'#ORACLE_BASE set from environment

member.__pga_aggregate_target=339738624

member.__sga_target=503316480

member.__shared_io_pool_size=0

member.__shared_pool_size=150994944

member.__streams_pool_size=0

*.audit_file_dest='/opt/oracle/admin/member/adump'

*.audit_trail='db'

*.compatible='11.2.0.0.0'

*.control_files='/opt/oracle/oradata/member/control01.ctl','/opt/oracle/flash_recovery_area/member/control02.ctl'

*.db_block_size=8192

*.db_domain=''

*.db_name='member'

*.db_recovery_file_dest='/opt/oracle/flash_recovery_area'

*.db_recovery_file_dest_size=4070572032

*.diagnostic_dest='/opt/oracle'

*.dispatchers='(PROTOCOL=TCP) (SERVICE=memberXDB)'

*.log_archive_format='%t_%s_%r.dbf'

*.memory_target=839909376

*.open_cursors=300

*.processes=150

*.remote_login_passwordfile='EXCLUSIVE'

*.undo_tablespace='UNDOTBS1'

#增加一下部分

*.db_unique_name='db1'

*.archive_lag_target=1800

*.fal_client='db1'

*.fal_server='db2'

*.log_archive_config='DG_CONFIG=(db1,db2)'

*.log_archive_dest_1='location=/opt/oracle/flash_recovery_area/ VALID_FOR=(all_logfiles,all_roles) db_unique_name=db1'

*.log_archive_dest_2='service=db2 lgwr async valid_for=(online_logfiles,primary_role) db_unique_name=db2'

*.log_archive_dest_state_1='enable'

*.log_archive_dest_state_2='enable'

*.log_archive_format='%t_%s_%r.dbf'

*.standby_file_management='auto'

*.db_file_name_convert='/opt/oracle/flash_recovery_area',' /opt/oracle/flash_recovery_area'

*.log_file_name_convert='/opt/oracle/flash_recovery_area ','/opt/oracle/flash_recovery_area'

並拷貝一個作爲備機db2的啓動參數文件

[oracle@db1 ~]$ cp /tmp/member.pfile /tmp/db2.pfile

[oracle@db1 ~]$ vim /tmp/db2.pfile

則將上面增加的部分修改爲

*.db_unique_name='db2'

*.archive_lag_target=1800

*.fal_client='db2'

*.fal_server='db1'

*.log_archive_config='DG_CONFIG=(db1,db2)'

*.log_archive_dest_1='location=/opt/oracle/flash_recovery_area/ VALID_FOR=(all_logfiles,all_roles) db_unique_name=db2'

*.log_archive_dest_2='service=db1 lgwr async valid_for=(online_logfiles,primary_role) db_unique_name=db1'

*.log_archive_dest_state_1='enable'

*.log_archive_dest_state_2='enable'

*.log_archive_format='%t_%s_%r.dbf'

*.standby_file_management='auto'

*.db_file_name_convert='/opt/oracle/flash_recovery_area',' /opt/oracle/flash_recovery_area'

*.log_file_name_convert='/opt/oracle/flash_recovery_area ','/opt/oracle/flash_recovery_area'

在db1上以修改過的/tmp/member.pfile啓動

(db1)SQL >shutdown immediate;

(db1)SQL> startup pfile='/tmp/member.pfile' nomount;

(db1)SQL> create spfile from pfile='/tmp/member.pfile';

(db1)SQL >shutdown immediate;

(db1)SQL> startup;

4.主庫密碼文件:

[1]存在密碼文件

[oracle@db1 dbs]$ ls $ORACLE_HOME/dbs

hc_DBUA0.dat hc_member.dat init.ora lkDB1 lkMEMBER orapwmember spfilemember.ora

看到上面有一個密碼文件orapwmember，在建庫的時候默認會創建一個

=================================================================================

[2]不存在密碼文件

如果沒有的話可以手動創建一個。

[oracle@db1 dbs]$ cd $ORACLE_HOME/dbs

[oracle@db1 dbs]$ orapwd file=orapwmember password=123456 entries=3

#注意以上需要根據SID名建立的，file=orapwSID

===================================================================================

5.修改監聽

[oracle@db1 dbs]$ cd $ORACLE_HOME/network/admin

[oracle@db1 admin]$ mv listener.ora listener.ora.default

[oracle@db1 admin]$ vim listener.ora

LISTENER =

(DESCRIPTION =

(ADDRESS = (PROTOCOL = TCP)(HOST = db1)(PORT = 1521))

)

SID_LIST_LISTENER =

(SID_LIST =

(SID_DESC =

(GLOBAL_DBNAME = member)

(ORACLE_HOME = /opt/oracle/product/11.2.0/db_1)

(SID_NAME = member)

)

#注意以上的HOST，就是HOSTNAME

[oracle@db1 admin]$ vim tnsnames.ora

MEMBER =

(DESCRIPTION =

(ADDRESS = (PROTOCOL = TCP)(HOST = localhost.localdomain)(PORT = 1521))

(CONNECT_DATA =

(SERVER = DEDICATED)

(SERVICE_NAME = member)

)

db1 =

(DESCRIPTION =

(ADDRESS_LIST =

(ADDRESS = (PROTOCOL = TCP)(HOST = db1)(PORT = 1521))

)

(CONNECT_DATA =

(SERVER = DEDICATED)

(SERVICE_NAME = db1)

)

db2 =

(DESCRIPTION =

(ADDRESS_LIST =

(ADDRESS = (PROTOCOL = TCP)(HOST = db2)(PORT = 1521))

)

(CONNECT_DATA =

(SERVER = DEDICATED)

(SERVICE_NAME = db2)

)

重啓下監聽

[oracle@db1 admin]$ lsnrctl stop

[oracle@db1 admin]$ lsnrctl start

6.複製監聽文件、參數文件、密碼文件到備庫

#注意一下我這邊$ORACLE_HOME

[oracle@db1 ~]$ echo $ORACLE_HOME

/opt/oracle/product/11.2.0/db_1

#傳輸啓動參數文件

[oracle@db1 ~]$ scp /tmp/db2.pfile db2:~

#傳輸密碼文件

[oracle@db1 ~]$ scp /opt/oracle/product/11.2.0/db_1/dbs/orapwmember db2:/opt/oracle/product/11.2.0/db_1/dbs/orapwmember

#傳輸監聽文件

[oracle@db1 ~]$ scp -r /opt/oracle/product/11.2.0/db_1/network/admin/{listener.ora,tnsnames.ora} db2:/opt/oracle/product/11.2.0/db_1/network/admin/

在db2上修改/opt/oracle/product/11.2.0/db_1/network/admin/listener.ora,將db1修改爲db2

[oracle@db2 ~]$ vim /opt/oracle/product/11.2.0/db_1/network/admin/listener.ora

LISTENER =

(DESCRIPTION =

(ADDRESS = (PROTOCOL = TCP)(HOST = db2)(PORT = 1521))

)

SID_LIST_LISTENER =

(SID_LIST =

(SID_DESC =

(GLOBAL_DBNAME = member)

(ORACLE_HOME = /opt/oracle/product/11.2.0/db_1)

(SID_NAME = member)

)

#只需要修改listener.ora即可，tnsnames.ora不需要動

7.主庫創建standby控制文件，我們這邊利用scp傳送全部文件

查看下控制文件的路徑

(db1)SQL> select name from v$controlfile;

NAME

--------------------------------------------------------------------------------

/opt/oracle/oradata/member/control01.ctl

/opt/oracle/flash_recovery_area/member/control02.ctl

(db1)SQL> shutdown immediate;

(db1)SQL> startup mount;

(db1)SQL> alter database create standby controlfile as '/opt/oracle/oradata/member/standby.ctl';

#在/opt/oracle/oradata/member/目錄下創建standby.ctl備機控制文件

8.複製主庫數據文件和日誌文件到備庫

[oracle@db1 ~]$ scp -r /opt/oracle/flash_recovery_area/ /opt/oracle/admin/ /opt/oracle/diag/ /opt/oracle/oradata/ db2:/opt/oracle

9.初始化備庫

在備機上使用standby的控制文件覆蓋原有的控制文件，覆蓋的路徑可以通過上一步查找控制文件的路徑瞭解到

[oracle@db2 ~]$ cp /opt/oracle/oradata/member/standby.ctl /opt/oracle/oradata/member/control01.ctl

[oracle@db2 ~]$ cp /opt/oracle/oradata/member/standby.ctl /opt/oracle/flash_recovery_area/member/control02.ctl

使用db2.pfile之前修改過的參數文件進行db2

(db2)SQL> startup pfile='/home/oracle/db2.pfile' nomount;

(db2)SQL> create spfile from pfile='/home/oracle/db2.pfile';

(db2)SQL> shutdown immediate;

(db2)SQL> startup nomount;

(db2)SQL> alter database mount standby database;

(db2)SQL> alter database open read only;

以下3種應用日誌的方法：（a和b選一）

a.開啓實時應用日誌，這樣在主庫插入立馬就可以在備機上查找到

(db2)SQL> alter database recover managed standby database using current logfile disconnect from session;

b.開啓redolog應用日誌，時間較長才能查詢到

(db2)SQL> alter database recover managed standby database disconnect from session;

c.停止應用redolog，只接受日誌，不重做

(db2)SQL> alter database recover managed standby database cancel;

至此，DataGuard搭建成功，在db1上創建表並插入數據，然後在db2上進行查詢就可以查到了。一開始做的時候總是查不到數據，最後發現是上面應用日誌的方法問題。

10.Dataguard測試：

查看Standby管理進程

(db1)SQL> select process,status from v$managed_standby;

PROCESS STATUS

--------- ------------

ARCH CONNECTED

ARCH CLOSING

LNS WRITING

(db2)SQL> select process,status from v$managed_standby;

PROCESS STATUS

--------- ------------

ARCH CLOSING

ARCH CONNECTED

MRP0 APPLYING_LOG

RFS IDLE

以上需要看到在主機上需要有LNS進程，在備機上需要RFS進程用來接收redo日誌，MRP0進程就是負責將日誌寫入數據庫中

在db1上進行切換日誌，然後在db2上查看日誌是否正常

(db1)SQL> select sequence#,applied from v$archived_log;

(db1)SQL> select max(sequence#) from v$archived_log;

MAX(SEQUENCE#)

--------------

(db2)SQL> select sequence#,applied from v$archived_log;

(db2)SQL> select max(sequence#) from v$archived_log;

MAX(SEQUENCE#)

--------------

(db1)SQL> alter system switch logfile;

(db1)SQL> select max(sequence#) from v$archived_log;

MAX(SEQUENCE#)

--------------

(db2)SQL> select max(sequence#) from v$archived_log;

MAX(SEQUENCE#)

--------------

從上面可以看出日誌切換成功，DataGuard正常運行

11.主備切換測試：

db1---primary/db2---standby ===》db2--primary/db1--standby

[oracle@db1 ~]$ lsnrctl stop

(db1)SQL> alter database commit to switchover to physical standby with session shutdown;

(db1)SQL> shutdown immediate;

(db1)SQL> startup mount;

(db1)SQL> alter database open read only;

(db1)SQL> alter database recover managed standby database using current logfile disconnect from session;

//在執行這條的時候，如果出現

ERROR at line 1:

ORA-01665: control file is not a standby control file

則是沒有執行alter database commit to switchover to physical standby with session shutdown;

[oracle@db1 ~]$ lsnrctl start

(db2)SQL> alter database commit to switchover to primary;

注意：

若出現ORA-16139: media recovery required，執行如下語句:

SQL> ALTER DATABASE RECOVER MANAGED STANDBY DATABASE DISCONNECT FROM SESSION;

SQL> alter database commit to switchover to primary;

如果出現,則可能是已打開了會話,加上with session shutdown強制關閉繪畫

ERROR at line 1:

ORA-01093: ALTER DATABASE CLOSE only permitted with no sessions connected

SQL> alter database commit to switchover to primary WITH SESSION SHUTDOWN;

(db2)SQL> shutdown immediate;

(db2)SQL> startup;

以上就是主備切換的流程

CentOS 5.11下Oracle 11G R2 Dataguard搭建

公司剛入職了一名 Java 中級開發，短短 4 行代碼居然湊齊了 3 個 bug！我哭了~~

公衆號5月C#/.NET熱文一覽

git 下載大陸鏡像地址

通過API自動增加Zabbix監控主機

Mariadb 10.3使用mariabackup備份進行主從搭建

Logstash將解析的日誌插入MySQL數據庫

kubeadm搭建高可用kubernetes 1.15.1

Nginx搶購限流配置

https://yachay.unat.edu.pe/blog/index.php?comment_area=format_blog&comment_component=blog&comment_co

linux以太網驅動總結