Oracle 增量修复DG同步

原创

布衣530

修改于 2025-01-26 16:02:24

3130

背景（报警不断的假期）

最近因灰度测试环境空间问题导致删除了部分未应用的归档，从而导致DG同步延迟。当时也在国庆假期，考虑到此为灰度测试环境备库供公司开发人员内部查询使用，就没有及时追平同步。假期结束后，发现归档差距有点大，于是考虑通过Oracle的增量备份恢复来修复DG的同步问题。

查看延迟

SQL> select value from v$dataguard_stats where name in ('apply lag');
VALUE
----------------------------------------------------------------
+5 04:33:48

Standby 查看SCN：

如果scn相同直接使用，如果有差异则选最小：10789509259

SQL> SELECT to_char(CURRENT_SCN) CURRENT_SCN from V$DATABASE;
CURRENT_SCN
----------------------------------------
10789509259

SQL> select to_char(min(checkpoint_change#)) min_scn from v$datafile_header;
MIN_SCN
----------------------------------------
10789509260

Primary:查看是否有新增数据文件：10789509259

SQL> col NAME for a50
SQL> select file# , NAME from v$datafile where creation_change# > =10789509259;
     FILE# NAME
---------- --------------------------------------------------
        31 /u01/oradata/two/two_dat12.dbf
        32 /u01/oradata/two/two_dat13.dbf

Primary:<SCN:10789509259>进行增量备份,新增的31、32数据文件同样需要备份

rman target /
run{
allocate channel c1 type disk;
allocate channel c2 type disk;
allocate channel c3 type disk;
allocate channel c4 type disk;
crosscheck archivelog all;
delete expired archivelog all; -- 删除所有已过期的归档日志文件
delete noprompt archivelog all completed before 'sysdate' ; -- 删除归档
-- 备份31、32数据文件（为断档后新增文件）
backup datafile 31,32 format '/home/oracle/rman/ForStandbyDat_%U' tag 'FORSTANDBY';
backup INCREMENTAL from scn 10789509259 database format '/home/oracle/rman/dat_incre_%U';
backup current controlfile for standby reuse format '/home/oracle/rman/standby_controlfile.ctl';
release channel c1;
release channel c2;
release channel c3;
release channel c4;
}

查看备份进度：

col opname format a35
col target_desc format a15
col perwork format a12
set lines 131
select sid,OPNAME,TARGET_DESC,sofar,TOTALWORK,trunc(sofar/totalwork*100,2)||'%' as perwork
from v$session_longops where sofar!=totalwork and totalwork!=0;

压缩并传输：

[oracle@two_db ~]$ tar -zcvf rman.tar.gz rman/
[oracle@two_db ~]$ scp -r rman.tar.gz two_standby:/home/oracle/

Standby 恢复

关库并恢复controlfile

SQL> shutdown immediate;
SQL> startup nomount
RMAN> restore standby controlfile from '/home/oracle/rman/standby_controlfile.ctl';
SQL> alter database mount;

注册catalog

RMAN> catalog start with '/home/oracle/rman';

RMAN恢复31、32号数据文件 – 新添加的数据文件必须restored到备库，无法通过备份恢复

RMAN> run
{
set newname for datafile 31 to '/u01/oradata/two/two_dat12.dbf'; 
set newname for datafile 32 to '/u01/oradata/two/two_dat13.dbf';
restore datafile 31;
restore datafile 32;
switch datafile all;
}

Standby开启恢复增量备份： – NOREDO意思就是不应用重做日志，仅应用增量备份

RMAN>run {
allocate channel c1 type disk;
allocate channel c2 type disk;
allocate channel c3 type disk;
allocate channel c4 type disk;
recover database noredo;
release channel c1;
release channel c2;
release channel c3;
release channel c4;
}

查看恢复进度

SQL> SELECT sid,serial#,CONTEXT,sofar,totalwork,round(sofar / totalwork * 100,2) "% Complete"
FROM v$session_longops
WHERE opname LIKE 'RMAN:%' AND opname NOT LIKE 'RMAN: aggregate%';

由于控制文件是从主库恢复过来的，所以需要重建standby log

SQL>alter system set STANDBY_FILE_MANAGEMENT=MANUAL;
SQL>alter database add  standby logfile
group 10 ('/u01/oradata/two/s_redo10.log') size 500M,
group 11 ('/u01/oradata/two/s_redo11.log') size 500M,
group 12 ('/u01/oradata/two/s_redo12.log') size 500M,
group 13 ('/u01/oradata/two/s_redo13.log') size 500M,
group 14 ('/u01/oradata/two/s_redo14.log') size 500M,
group 15 ('/u01/oradata/two/s_redo15.log') size 500M;
SQL>alter system set STANDBY_FILE_MANAGEMENT=auto;

Standby 恢复同步

Standby 开启日志同步进程：

SQL>alter database open read only;
SQL>ALTER DATABASE RECOVER MANAGED STANDBY DATABASE USING CURRENT LOGFILE DISCONNECT FROM SESSION;

Primary：激活：LOG_ARCHIVE_DEST_STATE_2

SQL> ALTER SYSTEM SET LOG_ARCHIVE_DEST_STATE_2=enable;
SQL> alter system switch logfile; -- 切换日志

Primary alert 日志：

******************************************************************
LGWR: Setting 'active' archival for destination LOG_ARCHIVE_DEST_2
******************************************************************
Wed Oct 16 17:34:02 2024
Archived Log entry 88573 added for thread 1 sequence 15431 ID 0xcf0caf5 dest 1:
Wed Oct 16 17:52:03 2024
Thread 1 advanced to log sequence 15433 (LGWR switch)
  Current log# 3 seq# 15433 mem# 0: /u01/oradata/two/redo03a.log
Wed Oct 16 17:52:03 2024
LNS: Standby redo logfile selected for thread 1 sequence 15433 for destination LOG_ARCHIVE_DEST_2
Wed Oct 16 17:52:03 2024
Archived Log entry 88576 added for thread 1 sequence 15432 ID 0xcf0caf5 dest 1:

Standby 查看同步情况：

SQL>  select value from v$dataguard_stats where name in ('apply lag');
VALUE
----------------------------------------------------------------
+00 00:00:00

– 至此整个修复完成

总结

操作文档下载：DG-增量恢复同步操作
增量备份：是基于以全量备份（0级备份）为基础的数据块的变化进行备份。从而降低备份成本：减少备份的数据量、存储的数据量以及备份数据传输的网络带宽。
本人经常使用的场景在异地灾备库扩盘，在异地保留一份备份，然后服务器在扩完盘再重做完系统，可能需要2-3天（因为是异地协调操作比较耗时）。2-3天的归档可能达到200GB+左右的归档，受限于专线带宽。增量备份就是不错的选择。（尤其异地灾备环境，非常受限于专线带宽）。

原创声明：本文系作者授权腾讯云开发者社区发表，未经许可，不得转载。

如有侵权，请联系 cloudcommunity@tencent.com 删除。

数据库