RMAN
recovery through RESETLOGS
Prior to Oracle 10g, the redo log files
generated after opening the database with RESETLOGS could not be used with the
backups taken before the RESETLOGS was performed.
Therefore, whenever a resetlogs was
done, it was important to take an immediate full database backup, since all
previous backups became invalid.
Just to recap, a RESETLOGS needs to be
performed when we
1) Do a point in time recovery
2) Recover a database using a backup of the control file
What does a RESETLOGS do?
·
Archives
the current online redo logs (if they are accessible) and then erases the
contents of the online redo logs and resets the log sequence number to 1.
·
Creates
the online redo log files if they do not currently exist.
·
Updates
all current datafiles and online redo logs and all subsequent archived redo
logs with a new
RESETLOGS
SCN
and time stamp
The Recovery through Resetlogs feature provides the
following benefits:
* There is no need to perform a full
backup after an incomplete recovery.
* There is no need to recreate a new
standby database after a failover operation.
* You can take incremental backups based
on full backups of a previous incarnation when you use RMAN.
* Block media recovery can restore
backups from parent incarnation backups and recover the corrupted blocks
through a RESETLOGS operation.
* You can use the newly generated logs
with an earlier incarnation of the database.
Important:
To perform recovery through RESETLOGS you must have all archived logs
generated after the most recent backup.
The new log_archive_format introduced in 10g
Oracle 10g introduces a new format
specification for archived log files. This new format avoids overwriting
archived redo log files with the same sequence number across incarnations.
SQL> show parameter log_archive_format
NAME TYPE VALUE
----------------------- ----------- ----------------
log_archive_format string %t_%s_%r.dbf
The format specification of the
log_archive_format string "%"r represents the resetlogs id. It will
ensure that a unique name is constructed for the archived redo log file during
RMAN restore and as well as restoring via SQL*plus auto recovery mode.
Note: The database would not start if
you remove the %r from the log archive format specification.
What is an incarnation?
A database incarnation is created whenever
you open the database with the RESETLOGS option.
The Current
incarnation is the one in which the database is running at present
The incarnation from which the current
incarnation branched after a ALTER DATABASE OPEN RESETLOGS was performed is called the Parent incarnation.
If
you want to go back to some SCN which is not part of the Current database
incarnation, we need to use the RESET DATABASE TO INCARNATION
command
as shown in the example below
The purpose of incarnations
An incarnation helps to identify redo
streams which have the same SCN, but occurred at different points in time. This
prevents the possibility of applying the wrong archive log file from a previous
incarnation which could corrupt the database.
Suppose we are at incarnation 1 and are
at SCN 100 in the database. I do a resetlogs and now the incarnation of the
database becomes 2. Suppose we do another resetlogs and it so happens that the
SCN at the time we did the resetlogs was also 100. Somewhere down the line
later we want to do a point in time
recovery and want to recover the database until SCN 100. How does Oracle know
which is the correct SCN and how far back in time should the database be rolled
back to and which archive logs need to be applied?
This is where incarnations come in and
we will see how we can set the database to a particular incarnation to enable
RMAN to do a point in time recovery.
Example
Let us now run the LIST INCARNATION OF DATABASE command from RMAN and this is the
output.
List of Database Incarnations
DB Key Inc Key DB
Name DB ID STATUS Reset SCN
Reset Time
------- ------- -------- ---------------- --- ----------
----------
1 1 TEST11G
916751459 PARENT 1
15-AUG-09
2 2 TEST11G
916751459 CURRENT
945184 18-OCT-11
What is the current SCN of the database?
SQL> select current_scn from v$database;
CURRENT_SCN
-----------
991395
Now, let us make a change in the
database which we will then try to undo by restoring and recovering the database to a point in
time before the media failure or as in this case, a wrong transaction has
happened.
SQL> conn sh/sh
Connected.
SQL> select count(*) from sales;
COUNT(*)
----------
918843
SQL> delete from sales where rownum < 1001;
1000 rows deleted.
SQL> commit;
Commit complete.
SQL> select count(*) from sales;
COUNT(*)
----------
917843
We now will roll the database back to an
SCN before the delete operation was performed.
SQL> conn / as sysdba
Connected.
SQL> shutdown immediate;
Database closed.
Database dismounted.
ORACLE instance shut down.
SQL> startup mount;
[oracle@kens-orasql-001 u01]$ rman target /
Recovery Manager: Release 11.2.0.1.0 - Production on Tue Oct 18
09:41:08 2011
Copyright (c) 1982, 2009, Oracle and/or its affiliates. All rights reserved.
connected to target database: TEST11G (DBID=916751459, not open)
RMAN> run {
2> set until scn 991395;
3> restore database;
4> recover database;
5> }
executing command: SET until clause
Starting restore at 18-OCT-11
using target database control file instead of recovery catalog
allocated channel: ORA_DISK_1
channel ORA_DISK_1: SID=63 device type=DISK
channel ORA_DISK_1: starting datafile backup set restore
channel ORA_DISK_1: specifying datafile(s) to restore from
backup set
channel ORA_DISK_1: restoring datafile 00001 to
/u01/oradata/TEST11G/system01.dbf
channel ORA_DISK_1: restoring datafile 00002 to
/u01/oradata/TEST11G/sysaux01.dbf
….
….
channel ORA_DISK_1: restored backup piece 1
channel ORA_DISK_1: restore complete, elapsed time: 00:00:15
Finished restore at 18-OCT-11
Starting recover at 18-OCT-11
using channel ORA_DISK_1
starting media recovery
media recovery complete, elapsed time: 00:00:00
Finished recover at 18-OCT-11
RMAN> alter database open resetlogs;
database opened
We can now see that the deleted rows
have been recovered and the number of rows in the table is now the same as
before the delete operation was performed.
SQL> select count(*) from sh.sales;
COUNT(*)
----------
918843
Let us check what the incarnation of the
database is.
RMAN> list incarnation of database;
using target database control file instead of recovery catalog
List of Database Incarnations
DB Key Inc Key DB
Name DB ID STATUS
Reset SCN Reset Time
------- ------- -------- ---------------- --- ----------
----------
1 1 TEST11G
916751459 PARENT 1
15-AUG-09
2 2 TEST11G
916751459 PARENT 945184
18-OCT-11
3 3
TEST11G 916751459 CURRENT 991396 18-OCT-11
The current incarnation of the database
is 3 and since we have recovered the database until SCN 991395, the RESET SCN
has been set to the SCN 991396.
The alert log of the database will also
log this resetlogs operation and we will see lines the ones shown below:
Incomplete
Recovery applied until change 991395 time 10/18/2011 09:39:19
Media Recovery
Complete (TEST11G)
Completed: alter
database recover if needed
start until change 991395
Tue Oct 18
09:42:14 2011
alter database
open resetlogs
Archived Log
entry 3 added for thread 1 sequence 5 ID 0x36a3e663 dest 1:
RESETLOGS after
incomplete recovery UNTIL CHANGE 991395
Setting recovery
target incarnation to 3
Tue Oct 18
09:42:15 2011
Assigning
activation ID 916768223 (0x36a4c5df)
Now let us make another delete from the
sales table.
SQL> delete from sh.sales where rownum < 10001;
10000 rows deleted.
SQL> commit;
Commit complete.
SQL> select count(*) from sh.sales;
COUNT(*)
----------
908843
Now, let us shutdown the database,
startup and mount it and try the same incomplete recovery which we tried
earlier until the SCN 991395
oracle@kens-orasql-001 u01]$ rman target /
Recovery Manager: Release 11.2.0.1.0 - Production on Tue Oct 18
09:46:29 2011
Copyright (c) 1982, 2009, Oracle and/or its affiliates. All rights reserved.
connected to target database: TEST11G (DBID=916751459, not open)
RMAN> run {
2> set until scn
991395;
3> restore database;
4> recover database;
5> }
executing command: SET until clause
Starting restore at 18-OCT-11
using target database control file instead of recovery catalog
RMAN-00571:
===========================================================
RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
RMAN-00571:
===========================================================
RMAN-03002: failure of restore command at 10/18/2011 09:46:55
RMAN-20208: UNTIL CHANGE is
before RESETLOGS change
We get the error RMAN-20208 because
since the current incarnation of the database is 3 and we trying to go to an
SCN before this incarnation.
So how do we go back to SCN 993195?
For this to happen, we need to change
the current incarnation (3) of the database to an older incarnation (2). The
SCN 993195 was present during the incarnation 2.
After resetting the incarnation to 2, we
now see that the restore is proceeding fine.
RMAN> reset database to incarnation 2;
database reset to incarnation 2
RMAN> run {
2> set until scn
991395;
3> restore database;
4> recover database;
5> }
executing command: SET until clause
Starting restore at 18-OCT-11
allocated channel: ORA_DISK_1
channel ORA_DISK_1: SID=63 device type=DISK
channel ORA_DISK_1: starting datafile backup set restore
channel ORA_DISK_1: specifying datafile(s) to restore from
backup set
channel ORA_DISK_1: restoring datafile 00001 to
/u01/oradata/TEST11G/system01.dbf
…
…
channel ORA_DISK_1: restore complete, elapsed time: 00:00:15
Finished restore at 18-OCT-11
Starting recover at 18-OCT-11
using channel ORA_DISK_1
starting media recovery
archived log for thread 1 with sequence 4 is already on disk as
file /u01/oradata/TEST11G/arch/1_4_764845989.arch
archived log for thread 1 with sequence 5 is already on disk as
file /u01/oradata/TEST11G/arch/1_5_764845989.arch
archived log file
name=/u01/oradata/TEST11G/arch/1_4_764845989.arch thread=1 sequence=4
archived log file
name=/u01/oradata/TEST11G/arch/1_5_764845989.arch thread=1 sequence=5
media recovery complete, elapsed time: 00:00:00
Finished recover at 18-OCT-11
Now we open the database with the ALTER
DATABASE OPEN RESETOGS command and see that a new incarnation key (4) has been
allocated to the database as the CURRENT incarnation and the previous
incarnation (3) has become an ORPHAN incarnation.
RMAN> list incarnation of database;
List of Database Incarnations
DB Key Inc Key DB
Name DB ID STATUS Reset SCN
Reset Time
------- ------- -------- ---------------- --- ----------
----------
1 1 TEST11G
916751459 PARENT
1 15-AUG-09
2 2 TEST11G
916751459 PARENT 945184
18-OCT-11
3 3 TEST11G
916751459 ORPHAN 991396
18-OCT-11
4 4 TEST11G
916751459 CURRENT
991396 18-OCT-11
Now let us do the same procedure once
again:
Delete some rows
Set the database to incarnation 2
Recover the database until the same SCN
as previous case – 991395
Open the database with resetlogs
Now list the incarnation of the database
RMAN> list incarnation of database;
List of Database Incarnations
DB Key Inc Key DB
Name DB ID STATUS Reset SCN
Reset Time
------- ------- -------- ---------------- --- ----------
----------
1 1 TEST11G
916751459 PARENT 1
15-AUG-09
2 2 TEST11G
916751459 PARENT 945184
18-OCT-11
4 4 TEST11G
916751459 ORPHAN 991396
18-OCT-11
5 5 TEST11G
916751459 CURRENT
991396 18-OCT-11
3 3 TEST11G
916751459 ORPHAN 991396
18-OCT-11
The incarnation key 5 now is the CURRENT
incarnation of the database, and incarnations 3 and 4 both have become ORPHAN.
The current SCN now is 991734
SQL> select current_scn from v$database;
CURRENT_SCN
-----------
991734
What happens if we change the structure
of the database? And we do not take any
new backup after this change was made, but try and recover using a backup
before we had changed the structure of the database. We add another datafile to
the USERS tablespace. So the current control file has knowledge of the fact
that the USERS tablespace now has not
one, but two datafiles.
We now try and do the same delete of
rows and try and go back to our old SCN 991395. At this point in time, the
control file had only knowledge of the fact that the USERS tablespace has one
data file and not two and the restore and recovery process will not try to do
anything with regards to the newly added datafile..
This is what we see in the alert log of
the database. Note that datafile 6 which is the new datafile added is being
taken offline and dropped because at SCN 991735, this datafile did not exist in
the database.
Wed Oct 19
10:03:42 2011
alter database
datafile 6 offline drop
Completed: alter
database datafile 6 offline drop
alter database
recover datafile list clear
Completed: alter
database recover datafile list clear
alter database
recover datafile list
1 , 2 , 3 , 4 , 5
Completed: alter
database recover datafile list
1 , 2 , 3 , 4 , 5
alter database
recover if needed
start until change 991735
Dictionary check
beginning
File #6 in the
controlfile not found in data dictionary.
Removing file
from controlfile.
data file 6:
'/u01/oradata/TEST11G/users02.dbf'
Dictionary check
complete
No comments:
Post a Comment