Oracle DataGuard Study之--DataGuard FailOver案例

23dq1

Oracle DG（Dataguard）是目前比较常见的数据库HA配置策略。通过实现Physical Standby和Logical Standby，可以实现数据冗余容错机制。防止在主库出现严重故障，不能支持服务的时候，没有快速的后备支持环境。

在DG中，switchover和failover是两个重要的概念，也是DG实现的核心。两者共同点都是Primary和Standby角色切换，差异在于Planned和UnPlanned之分。Switchover关键点在于Planned，这个切换动作是在运维机构规划范围内的动作。比如，进行定期系统软硬件升级、设备维修等动作。而Failover是真正出现严重系统故障，如数据库宕机、软硬件故障导致的Primary不能支持服务，从而进行的切换动作。

根据不同的DG配置，switchover和failover也是有差异的。理论上，Switchover是不会造成数据丢失的，Primary在切换之后也是在DG配置环境中，作为Standby存在的。但是Failover则不同，除了运行在最大保护（Maximum Protection）模式下，Primary突发的故障可能引起一部分Redo Log不能及时的传递到Standby端，切换之后很可能有数据损失的情况。更重要的是，Primary端在发生Failover之后，是不能够直接加入回DG配置的！也就是说，Failover之后，Primary实际上就是被“抛出”了DG环境。

那么，有什么方法实现Primary回到原有的环境呢？这个问题的困难在于保持Primary和Standby一致。在正常情况下，Primary和Standby之间是关联同步的，即使发生了Switchover，也在可控情况下。Failover过程中有数据的缺失，还有Primary修复问题。在目前流行版本（11g）中，有三个方法：

ü  环境重建：一种最简单的方法就是直接删除原来的Primary库，引用DG重建方法，重新搭建Standby端；

ü  RMAN备份恢复：如果Primary端保留过一份Failover之前的备份，则可以强制原来的Primary端恢复到进行Failover的时间点，之后作为Standby接收当前Primary的redo log传递，应用后可以跟上进度；

ü  Flashback Database恢复：Flashback技术是作为传统备份还原技术的补充，提供了更加便捷的恢复策略。使用flashback，可以将数据库恢复到failover之前的时间点。之后的过程和RMAN备份恢复策略相同；

案例分析：

一、在主库端模拟数据库意外宕机

1 2 3 4 5 6 7

scott@bjdb>conn /as sysdba Connected. sys@bjdb>alter system switch logfile; System altered. sys@bjdb>shutdown abort ORACLE instance shut down.

二、在备库端
1、查看切换信息

1 2 3 4 5

sys@shdb>select name,database_role,switchover_status from v$database; NAME      DATABASE_ROLE    SWITCHOVER_STATUS --------- ---------------- -------------------- TESTDB12  PHYSICAL STANDBY NOT ALLOWED 可以看到此时备库处于无法切换状态

2、直接切换

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

sys@shdb>alter database commit to switchover to primary; alert_log:（告警日志） Fatal NI connect error 12514, connecting to: (DESCRIPTION=(ADDRESS_LIST=(ADDRESS=(PROTOCOL=TCP)(HOST=shsrv)(PORT=1521)))(CONNECT_DATA=(SERVICE_NAME=shdb)(CID=(PROGRAM=oracle)(HOST=bjsrv)(USER=oracle))))   VERSION INFORMATION:         TNS for Linux: Version 11.2.0.3.0 - Production         TCP/IP NT Protocol Adapter for Linux: Version 11.2.0.3.0 - Production   Time: 04-MAR-2015 21:25:13   Tracing not turned on.   Tns error struct:     ns main err code: 12564       TNS-12564: TNS:connection refused     ns secondary err code: 0     nt main err code: 0     nt secondary err code: 0     nt OS err code: 0 Error 12514 received logging on to the standby FAL[client, MRP0]: Error 12514 connecting to shdb for fetching gap sequence Wed Mar 04 21:26:00 2015 alter database commit to switchover to primary ALTER DATABASE SWITCHOVER TO PRIMARY (TestDB12) Maximum wait for role transition is 15 minutes. Switchover: Media recovery is still active Database not available for switchover   End-Of-REDO archived log file has not been recovered Database not available for switchover   End-Of-REDO archived log file has not been recovered Database not available for switchover

3、关闭standby MPR进程

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35

sys@shdb>ALTER DATABASE RECOVER  managed standby database finish; ALTER DATABASE RECOVER  managed standby database finish   Terminal Recovery: request posted (TestDB12) Wed Mar 04 21:34:34 2015 Begin: Standby Redo Logfile archival End: Standby Redo Logfile archival Terminal Recovery timestamp is '03/04/2015 21:34:34' Terminal Recovery: applying standby redo logs. Terminal Recovery: thread 1 seq# 34 redo required Media Recovery Waiting for thread 1 sequence 34 Terminal Recovery: End-Of-Redo log allocation Terminal Recovery: standby redo logfile 4 created '/dsk4/arch_bj/arch_1_0_820054583.log' This standby redo logfile is being created as part of the failover operation.  This standby redo logfile should be deleted after the switchover to primary operation completes. Media Recovery Log /dsk4/arch_bj/arch_1_0_820054583.log Terminal Recovery: log 4 reserved for thread 1 sequence 34 Recovery of Online Redo Log: Thread 1 Group 4 Seq 34 Reading mem 0   Mem# 0: /dsk4/arch_bj/arch_1_0_820054583.log Identified End-Of-Redo (failover) for thread 1 sequence 34 at SCN 0xffff.ffffffff Incomplete Recovery applied until change 1234252 time 03/04/2015 21:23:43 MRP0: Media Recovery Complete (TestDB12) Terminal Recovery: successful completion Wed Mar 04 21:34:35 2015 ARCH: Archival stopped, error occurred. Will continue retrying ORACLE Instance TestDB12 - Archival Error ORA-16014: log 4 sequence# 34 not archived, no available destinations ORA-00312: online log 4 thread 1: '/dsk4/arch_bj/arch_1_0_820054583.log' Forcing ARSCN to IRSCN for TR 0:1234252 Attempt to set limbo arscn 0:1234252 irscn 0:1234252 Resetting standby activation ID 2865247982 (0xaac836ee) MRP0: Background Media Recovery process shutdown (TestDB12) Terminal Recovery: completion detected (TestDB12) Completed: ALTER DATABASE RECOVER  managed standby database finish

4、切换数据库到Primary

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39

sys@shdb>select status from v$instance; STATUS ------------ OPEN sys@shdb>select name,database_role,switchover_status from v$database; NAME      DATABASE_ROLE    SWITCHOVER_STATUS --------- ---------------- -------------------- TESTDB12  PHYSICAL STANDBY TO PRIMARY sys@shdb>alter database commit to switchover to primary; Database altered. sys@shdb>alter database open; Database altered. 告警日志： alter database commit to switchover to primary ALTER DATABASE SWITCHOVER TO PRIMARY (TestDB12) Maximum wait for role transition is 15 minutes. All dispatchers and shared servers shutdown CLOSE: killing server sessions. CLOSE: all sessions shutdown successfully. Wed Mar 04 21:35:47 2015 SMON: disabling cache recovery Backup controlfile written to trace file /u01/app/oracle/diag/rdbms/bjdb/TestDB12/trace/TestDB12_ora_3146.trc Standby terminal recovery start SCN: 1234251 RESETLOGS after incomplete recovery UNTIL CHANGE 1234252 Online log /dsk2/oradata/bjdb/redo01b.log: Thread 1 Group 1 was previously cleared Online log /dsk1/oradata/bjdb/redo01a.log: Thread 1 Group 1 was previously cleared Online log /dsk2/oradata/bjdb/redo02b.log: Thread 1 Group 2 was previously cleared Online log /dsk1/oradata/bjdb/redo02a.log: Thread 1 Group 2 was previously cleared Online log /dsk2/oradata/bjdb/redo03b.log: Thread 1 Group 3 was previously cleared Online log /dsk1/oradata/bjdb/redo03a.log: Thread 1 Group 3 was previously cleared Standby became primary SCN: 1234250 Wed Mar 04 21:35:47 2015 Setting recovery target incarnation to 3 AUDIT_TRAIL initialization parameter is changed back to its original value as specified in the parameter file. Switchover: Complete - Database mounted as primary Completed: alter database commit to switchover to primary

三、原主库修复后，开机

1 2 3 4 5 6 7 8 9 10 11 12 13 14

sys@bjdb>startup ORACLE instance started. Total System Global Area  442601472 bytes Fixed Size                  2229184 bytes Variable Size             281021504 bytes Database Buffers          155189248 bytes Redo Buffers                4161536 bytes Database mounted. Database opened. sys@bjdb>select name,database_role,switchover_status from v$database; NAME      DATABASE_ROLE    SWITCHOVER_STATUS --------- ---------------- -------------------- TESTDB12  PRIMARY          FAILED DESTINATION

    现在原来的主库被修复后，整个DataGuara架构已经被破坏了，所以必须把原来的主库构建成新的备库，重新恢复DataGuard的环境。

四、重新构建DataGuard

1	sys@bjdb>select name,database_role from v$database;

NAME                                               DATABASE_ROLE

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

TESTDB12                                           PHYSICAL STANDBY