Oracle ADG + Keepalived 切換演練

客戶的一套生產環境採用的架構是Oracle ADG + Keepalived，近期須要進行切換演練，要求我這邊保障。ADG自己切換倒沒啥可說的，但引入keepalived軟件，就須要提早研究下這個架構。其實看了下環境配置，總體思路也很是簡單，說白了就是利用keepalived軟件引入一個VIP，應用側只需配置鏈接這個VIP便可。
依據當前生產環境架構模擬了一套本身的測試環境。

• 1.Keepalived相關配置
• 2.ADG手工切換步驟
• 3.VIP和監聽的關係

1.Keepalived相關配置

關於Keepalived軟件的配置和編譯安裝，能夠參考以前《MySQL主主+Keepalived架構安裝部署》中Keepalived安裝部署章節。 除了利用keepalived軟件引入一個VIP，還有一些配置和腳本，脫敏以下：

--------------------------------------------------------
--節點1（192.168.1.124）keepalived.conf文件內容：
--------------------------------------------------------
[root@test04 ~]# cat /etc/keepalived/keepalived.conf
! Configuration File for keepalived

vrrp_script chk_dg_stats {  
    script "/etc/keepalived/check_dataguard.sh" 
    interval 2
    weight -5
    fall 2  
    rise 1  
}
       
vrrp_instance VI_1 {
    state MASTER    
    interface eth0 
    mcast_src_ip 192.168.1.124
    virtual_router_id 131 
    priority 101 
    inopreempt
    advert_int 1         
    authentication {   
        auth_type PASS 
        auth_pass 888888   
    }
    virtual_ipaddress {    
        192.168.1.131
    }
      
    track_script {               
       chk_dg_stats             
    }
}

--------------------------------------------------------
--節點2（192.168.1.125）keepalived.conf文件內容：
--------------------------------------------------------
[root@test05 ~]# cat /etc/keepalived/keepalived.conf
! Configuration File for keepalived

vrrp_script chk_dg_stats {  
    script "/etc/keepalived/check_dataguard.sh" 
    interval 2
    weight -5
    fall 2  
    rise 1  
}
       
vrrp_instance VI_1 {
    state BACKUP
    interface eth0 
    mcast_src_ip 192.168.1.125
    virtual_router_id 131 
    priority 99 
    inopreempt
    advert_int 1         
    authentication {   
        auth_type PASS 
        auth_pass 888888   
    }
    virtual_ipaddress {    
        192.168.1.131
    }
      
    track_script {               
       chk_dg_stats             
    }
}

--------------------------------------------------------
--全部節點配置腳本check_dataguard.sh，並確認具備x執行權限：
--------------------------------------------------------
# cat /etc/keepalived/check_dataguard.sh 
#!/bin/bash
dbstats=`ps -ef | grep ora_smon | grep -v grep | wc -l`
dgstats=`ps -ef | grep ora_mrp | grep -v grep | wc -l`

if [ "${dbstats}" -eq 0  ]; then
/etc/init.d/keepalived stop
elif [[ "${dbstats}" -gt 0 ]] && [[ "${dgstats}" -gt 0 ]]; then
/etc/init.d/keepalived stop
fi 

說明：腳本check_dataguard.sh主要經過對ora_smon和ora_mrp進程的監控，判斷哪一種場景下該關閉keepalived服務：
場景1:當不存在ora_smon進程時（數據庫實例Crash）；
場景2:存在ora_smon進程同時存在ora_mrp進程時（已啓動mrp進程的備庫）。

--添加x執行權限：
chmod u+x /etc/keepalived/check_dataguard.sh
[root@test04 ~]# ls -l /etc/keepalived/check_dataguard.sh
-rwxr--r--. 1 root root 282 Jul 14 22:35 /etc/keepalived/check_dataguard.sh
[root@test05 ~]# ls -l /etc/keepalived/check_dataguard.sh
-rwxr--r--. 1 root root 281 Jul 14 22:36 /etc/keepalived/check_dataguard.sh

2.ADG手工切換步驟

1）在switchover正式切換前先在主庫上手工切換幾回日誌，確認DG備庫同步正常： 
--PRIMARY（主庫192.168.1.124）切換幾回日誌:
SQL>
alter system switch logfile; 
alter system switch logfile; 
alter system switch logfile; 
--Standby （備庫192.168.1.125）需確認同步正常沒有延遲:
SQL> 
select * from v$dataguard_stats; 
2）主庫切換爲備庫
-- 在PRIMARY（主庫192.168.1.124）查詢，確承認切換爲備庫：
select OPEN_MODE, DATABASE_ROLE, SWITCHOVER_STATUS, FORCE_LOGGING, DATAGUARD_BROKER, GUARD_STATUS from v$database; 
-- 在PRIMARY（主庫192.168.1.124）操做，切換爲備庫：
ALTER DATABASE COMMIT TO SWITCHOVER TO STANDBY WITH SESSION SHUTDOWN;
3）備庫切換爲主庫
-- 在Standby（備庫192.168.1.125）查詢，確承認切換爲主庫：
select OPEN_MODE, DATABASE_ROLE, SWITCHOVER_STATUS, FORCE_LOGGING, DATAGUARD_BROKER, GUARD_STATUS from v$database; 
-- 在Standby（備庫192.168.1.125）操做，切換爲主庫（根據SWITCHOVER_STATUS值確認用下面哪一個命令）：
ALTER DATABASE COMMIT TO SWITCHOVER TO PRIMARY;
ALTER DATABASE COMMIT TO SWITCHOVER TO PRIMARY WITH SESSION SHUTDOWN;

4）新主庫open，新備庫啓動並開啓MRP，新主庫啓動keepalived服務
--NEW PRIMARY（新主庫192.168.1.125）數據庫從mount啓動到open狀態:
ALTER DATABASE OPEN;
--NEW STANDBY（新備庫192.168.1.124）數據庫startup啓動，開啓DG日誌應用:
STARTUP
RECOVER MANAGED STANDBY DATABASE USING CURRENT LOGFILE DISCONNECT;
--確認NEW STANDBY（新備庫192.168.1.124）DG同步正常，沒有延遲：
SQL> 
select * from v$dataguard_stats; 

5) 新主庫啓動keepalived服務
--NEW PRIMARY（新主庫192.168.1.125）OS層root用戶啓動keepalived服務：
# /etc/init.d/keepalived start

注意：當演練結束後，若須要switchover主備再次切換，只須要按上面規範步驟重複操做便可（注意主備角色的轉換）。

3.VIP和監聽的關係

源於最先的一次面試，兩個節點的RAC，節點1主機Crash，此時應用經過節點1的VIP是否能夠鏈接到數據庫？爲何？ 咱們都知道節點1主機Crash，其VIP會自動漂移節點2，ping這個IP也是通的，可是經過其鏈接數據庫卻不行！會報一個沒有監聽（ORA-12541: TNS:no listener）的錯誤。 具體可參考：

RAC 某節點不可用時，對應VIP是否可用
那這裏的環境，一樣是VIP的設置，爲什麼卻能夠經過VIP（192.168.1.131）鏈接呢？

[oracle@test03 ~]$ sqlplus sys/oracle@192.168.1.131/demo as sysdba

SQL*Plus: Release 11.2.0.4.0 Production on Tue Jul 14 23:45:23 2020

Copyright (c) 1982, 2013, Oracle.  All rights reserved.


Connected to:
Oracle Database 11g Enterprise Edition Release 11.2.0.4.0 - 64bit Production
With the Partitioning, OLAP, Data Mining and Real Application Testing options

SQL>

實際驗證，是由於這裏主備庫的監聽配置統一都是主機名:

[oracle@test04 admin]$ cat listener.ora
# listener.ora Network Configuration File: /u01/app/oracle/product/11.2.0/dbhome_1/network/admin/listener.ora
# Generated by Oracle configuration tools.

LISTENER =
  (DESCRIPTION_LIST =
    (DESCRIPTION =
      (ADDRESS = (PROTOCOL = IPC)(KEY = EXTPROC1521))
      (ADDRESS = (PROTOCOL = TCP)(HOST = test04)(PORT = 1521))
    )
  )

ADR_BASE_LISTENER = /u01/app/oracle

[oracle@test05 admin]$ cat listener.ora 
# listener.ora Network Configuration File: /u01/app/oracle/product/11.2.0/dbhome_1/network/admin/listener.ora
# Generated by Oracle configuration tools.

LISTENER =
  (DESCRIPTION_LIST =
    (DESCRIPTION =
      (ADDRESS = (PROTOCOL = IPC)(KEY = EXTPROC1521))
      (ADDRESS = (PROTOCOL = TCP)(HOST = test05)(PORT = 1521))
    )
  )

ADR_BASE_LISTENER = /u01/app/oracle

SID_LIST_LISTENER =
  (SID_LIST =
    (SID_DESC =
      (GLOBAL_DBNAME = jingyus)
      (ORACLE_HOME = /u01/app/oracle/product/11.2.0/dbhome_1)
      (SID_NAME = jingyu)
    )
  )

若是將主機名修改成具體的IP地址，則測試一樣會報錯（ORA-12541: TNS:no listener）。