DM8 uses different databases to archive and recover after multiple failures

This chapter is an introduction. DM8 The database recovers after many failures , The latest database is not backed up , How to use the previous backup file to restore to the latest state of the database .

Use different archives for recovery after multiple failures

01

Backup database

Want to restore and recover in case of database failure , There must be backup files , Here is a most original database backup file , have access to DMRMAN Backup , It can also be an online backup file . Take online backup as an example .

Query the magic number of the database ：

SQL> select db_magic;

Line number     DB_MAGIC            

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

1          2098604336

Backup database ：

SQL> backup database full backupset 'DMAMENGFULL_DMRMAN0716';

02

Manufacturing data , Generate archive logs 1

Use DMHR Connect to the login database , Use the following command to create a table t_table1, This process generates archive logs 1.

SQL> create table dmhr.t_table1 as select * from dmhr.employee;

03

Analog fault , Delete data file

Analog fault , perform rm DMHR.DBF Delete data file

04

Restore the database with backup files and archived logs

To the latest state （ Status before failure ）

Stop database , Use DMRMAN Restore 、 Restore the database and update the magic number of the database . The following three steps are a typical way to restore and recover a single database failure using archived files .

RMAN> restore database '/dm8/data/DAMENG/dm.ini' from backupset 'DMAMENGFULL_DMRMAN0716';RMAN> recover database  '/dm8/data/DAMENG/dm.ini' with archivedir '/dm8/arch';RMAN> recover database  '/dm8/data/DAMENG/dm.ini' update db_magic;
05

 Open database , Manufacturing data , Generate 

Archive logs after database recovery 2

Open database , Query database DMHR Next t_table1 There is , The data is correct , It indicates that the database has been restored to the latest state . Execute the following command to view DB_MAGIC, You can see DB_MAGIC Has changed .

SQL> select db_magic;

Line number     DB_MAGIC            

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

1          813815328

Use DMHR Connect to the login database , establish t_table2 surface , Generate archive logs after database recovery 2.

SQL> create table dmhr.t_table2 as select * from dmhr.employee;

06

Analog fault , Delete data file

Analog fault , perform rm DMHR.DBF Delete data file

07

Follow the steps 1 Backup files and twice

Archive log restore and restore database

Because we didn't back up after the first restore , So it can only be based on steps 1 Restore the backup file generated in , Of course, if we are in step after the first failure 4 After recovery, the backup was done normally , The recovery at this time will also be very simple , Direct use procedure 4 Single fault recovery operation is ok . But because there are only steps 1 Backup files for , Experienced two failures in the middle , So we are going to talk about cross database （ Magic number across databases ） recovery , The recovery steps are as follows ：

① Use steps 1 Restore the backup file of

Execute the following command to use the steps 1 Restore the database to the state at the time of backup ：

RMAN> restore database '/dm8/data/DAMENG/dm.ini' from backupset 'DMAMENGFULL_DMRMAN0716';

② Use steps 2 Archive logs generated 1 Recover to the first failure state

Execute the following command , Appoint with archivedir Parameter use archive recovery , When recovering from an archive , By default, the target library's db_magic, At this point, the target library db_magic The value is replaced by step 1 Backup set source library db_magic, So steps are used 2 Archive logs generated .

RMAN> recover database  '/dm8/data/DAMENG/dm.ini' with archivedir '/dm8/arch';

③ Use steps 5 Archive logs generated 2 Restore to the latest state

Execute the following command , Use use db_magic Parameters , Appoint db_magic The value is to be restored to the archive db_magic value （ That is, use steps 5 Of db_magic, Use steps 5 Archive recovery generated ）, Restore the data to step 6 Status before failure .

RMAN> recover database  '/dm8/data/DAMENG/dm.ini' with archivedir '/dm8/arch' use db_magic 813815328;

④ Update database magic number

Execute the following command to update the magic number of the database db_magic Information .

RMAN> recover database  '/dm8/data/DAMENG/dm.ini' update db_magic;

here , Database recovery complete , Open database , You can see dmhr Next t_table2 The table exists and the data is accurate . If you query the magic number of the database , You can see db_magic Updated .

Use dmrachk Tool View db_magic Information

Actually in the database of the current network , Unable to start after multiple database failures , We may not know the magic number information in the database every time . have access to dmrachk The tool checks db_magic Information . perform dmrachk help You can check the usage of this tool .

Execute the following command to view the archive log information , The red part is the magic number of the database .

dmrachk arch_fil=ARCHIVE_LOCAL1_0x51FA5B48[0]_2020-07-16_14-49-56.log

Checked the archive log db_magic Information , Then execute the 7 Step by step ③ Use use magic You can specify the archive log file recovery of the magic number of the database .

good , The above is the content of this sharing .

summary ：

In case of multiple failure recovery , You can specify use db_magic Parameter recover database multiple times , Until it is restored to the latest state .

When using archive recovery, attention should be paid to ensuring the continuity of archive log files , If the database closes the archive in the middle and then reopens the archive , Or the archive log file is lost, causing the archive log to be discontinuous , The database cannot be restored to the latest state .