Oracle Online backup limitations

Hi,
We are using oracle 10g for SAP.
1. In which situation oracle Online consistent backup is not useful.
2. In a situation all online relog groups are corrupted. Is it possible recovery with online consistent backup ?
3. In which recovery situation offline backup is good than online backup?
Regards,
Nani

Hi,
2. In a situation all online relog groups are corrupted.
Is it possible recovery with online consistent backup ?
Online Consistent backup only contains the offline redo log files created during a backup run. One can not get any online redo log group members from such online consistent backup. If all online redo logs are corrupted/loss then in such situation instance recovery is not possible and only incomplete partial /point-in-time recovery is possible. If the mirrored online redo log files is available then its possible to perform instance recovery.
Please refer "Recovering from Current Online Redo Log Missing" && "Recovering from One Inactive Online Redo Log Missing" topic of this useful document [Oracle Restore and Recovery|http://www.sdn.sap.com/irj/scn/go/portal/prtroot/docs/library/uuid/97a4f439-0d01-0010-d488-d29b45e2f5ab?quicklink=index&overridelayout=true].
I would refer this useful document [Oracle Restore and Recovery|http://www.sdn.sap.com/irj/scn/go/portal/prtroot/docs/library/uuid/97a4f439-0d01-0010-d488-d29b45e2f5ab?quicklink=index&overridelayout=true] (specially its last part) to get more information about different scenario of Restore &recovery in Oracle.
Regards,
Bhavik G. Shroff

Similar Messages

Oracle online backup failed

Dear all,
I am scheduling a online backup for ECC 6.0 , oracle 10.2 to target NAS device. I have mentioned the error log.
BR0252E Function _mkdir() failed for '
<IPaddress>\backup\bebfoyky' at location BrDirCreate-1
BR0253E errno 13: Permission denied
I have created sapservice<sid> user in the NAS device. Anything i need to check more. Could anyone have any idea... please help me.
Thanks
Raj

Dear Juan,
Yes , both the users belongs to administrator group.
@ Rishi - The Ip address which i have mentioned in my earlier mess is the IP address of the network device.
It was working fine with another device, after changing it to the new target device i am facing this issue.
I have changed the target network path of the NAS in the init<sid>.sap, also i have restarted the instance as well.
Do i need to reboot the entire server to make the changes effect ?
Kindly suggest me.
BR0252E Function _mkdir() failed for '
<IP address of the NAS>\backup\bebfoyky' at location BrDirCreate-1
BR0253E errno 13: Permission denied
Thanks
Raj

Oracle online backup

Is it possible to play future archive logs into a hot backup in the same manner that you can with a cold backup.

Recover ...until cancel yields the error below: and the standby could actually be better described as a staging area for application development.
QL> recover database until cancel;
ORA-00283: recovery session canceled due to errors
ORA-01610: recovery using the BACKUP CONTROLFILE option must be done
SQL> recover database using backup controlfile until cancel;
ORA-00279: change 169829095 generated at 12/26/2006 23:38:35 needed for thread
1
ORA-00289: suggestion : D:\ORACLE\ADMIN\EMRDB\ARCH\ARC057892785102977.001
ORA-00280: change 169829095 for thread 1 is in sequence #2977
Specify log: {<RET>=suggested | filename | AUTO | CANCEL}
D:\oracle\Data\EMRDB\REDO1A.LOG
ORA-00326: log begins at change 171504873, need earlier change 169829095
ORA-00334: archived log: 'D:\ORACLE\DATA\EMRDB\REDO1A.LOG'
ORA-01547: warning: RECOVER succeeded but OPEN RESETLOGS would get error below
ORA-01195: online backup of file 1 needs more recovery to be consistent
ORA-01110: data file 1: 'D:\ORACLE\DATA\EMRDB\SYSTEM.DB'

Oracle Online Backup through EMC Networker - RMAN Script

Hi,
For taking online oracle backup through EMC Networker, we configure the RMAN script to connect to the target database, create the media channels, backup format etc. The password for the sys or rman user entered in the RMAN script is in clear-text format.
Due to security reasons, I don't want the passwords to be entered in the RMAN script. Is there any option for configuring RMAN script without mentioning the username/password to connect to the target database ?
Thanks & Regards,
Shreejesh.

When you have to connect remotely to RMAN, you cannot hide the password, you can only define very restricted access for the script itself. Remote OS authentication will not work, because the target db user needs SYSDBA privilege, which cannot be granted to an externally identified user.
Werner

Online backups on BIW systems, every update is logged?

Some years ago I was at an SAP BIW installation, they performed Oracle online backups dayly.
For some reason, they had to restore and recover DB, and after using an online backup as a restore basis, some indexes were not usable, because they had been updated without archive log information.
Finally, the easiest way for the system administator was using an offline backup.
I got to know there were some updates (not critical) not being logged in the DB, so online backups, as well as standby DB would not be as useful as on OLTP systems.
Does somebody know if this is or this is not the same nowadays? Is everything logged on a BW system? Is that a tunable configuration for some indexes? Is an online backup really useful after restoring and recovering?
Thanks very much for your experience and answers !!
Eva

Hello Eva,
> For some reason, they had to restore and recover DB, and after using an online backup as a restore basis, some indexes were not usable, because they had been updated without archive log information.
This behaviour is absolutely normal.
In a BW system the indexes are created / rebuild with the NOLOGGING clause, because of some performance reasons.
> Finally, the easiest way for the system administator was using an offline backup.
The easiest yes - but concurrently the worst one. He can identify the "new" indexes since the last online backup and only recreate this ones.
> Does somebody know if this is or this is not the same nowadays?
Yes the NOLOGGING clause is still used nowadays.
> Is everything logged on a BW system?Is that a tunable configuration for some indexes?
You can "deactivate" the NOLOGGING clause.. but then you can have some performance impacts in your BW load processes.
Check sapnote #442763 for the customizing.
> Is an online backup really useful after restoring and recovering?
Yes, of course .. the ONLINE backup is the preferred method.
If you recover your database with the BR*Tools .. your unusable indexes are automatically rebuild.
Check sapnote #849485 for that feature.
Regards
Stefan

Online backup of the Oracle Collaboration Suite Database automatically ever

Hi everybody!
How to perform an online backup of the Oracle Collaboration Suite Database automatically every half-hour (every 10 minutes)?
Thanks a lot for your helps.

Hi,
Take a look at the Backup and Recovery Basics book in the Administration section of the Database library. This will explain how ArchiveLog works.
richard

Oracle Restore from Online backup without any archive logs

Hi,
May be a dumb question. If I have a good online backup (say it took 2 hours to do that. And there is database activity while backup is going on), and lost all the archive logs happened after the online backup, is it possible at all to do the restore using that backup? complete or incomplete? and bring back the database to normal operation mode? Some details on this.
Thanks.

Let us see the reasoning behind this.
Database:WORLDDB
WORLDDB configuration:-
TBSP_WDB
->TBSP_WDB_01.dbf
->TBSP_WDB_02.dbf
SYSTEM
->SYSTEM.dbf
USERS
and so on for the tablespaces.
CASE 1:-
User managed backup.
You issue
ALTER TABLESPACE TBSP_WDB BEGIN BACKUP;
DB keeps working all transactions are recorded in the redo stream, the scn information is not updated in the file header.
Also note that the other tablespaces were been continuously worked on hence their scn numbers are a moving target.
http://download-east.oracle.com/docs/cd/B10501_01/server.920/a96572/osbackups.htm#10012
So if you begin tablespace TBSP_WDB backup at time t1 and by the time you are done with your backups you are at time t2, you will need all of the archive log's between time t1 and t2.
CASE 2:-
If we do with rman, rman takes care of the file header update information and goes on its merry way, those details are hidden from the end user. If one of the tablespace is fractured, updated while backups are going on , rman knows about it and will re-read the affected blocks in question.
My thinking would be that you might need archive logs, unless it is a cold backup, with switch logfiles in between.
If you do hot backups you need archivelogs is my thought. The number of archivelogs required will be decreased considerably through rman.
http://download-east.oracle.com/docs/cd/B10501_01/server.920/a96566/rcmconc1.htm#458821
Summary:-
Either way archivelogs are needed.

How does recovery work after an online backup

Hello,
While trying to conceptually understand how backup and recovery works, I came accross a question concerning hot (online) backup.
This is a conceptual question (I am trying to understand how things work), it is not a "how should I proceed/ what should I do step by step" question.
As far as I understand, an online backup of a tablespace can be performed by copying the OS files making up a tablespace while the database is up and being used (i.e. transactions are modifying data in the database). Before the copying of the OS files starts, the Oracle RDMS must be notified that an online backup is being taken via "ALTER...BEGIN BACKUP" (such that some additional information is written to the Redo Log, which may be required for subsequent recovery using the online backup). During recovery the Oracle RDBMS uses the copies of the OS files together with the online and archived redo logs in order to reconstruct all committed transactions and it further uses the UNDO tablespace to rollback open (uncommitted) transactions.
Thinking about this, it seems to me, that in order for this to work in all possible scenarios the undo information from the time the backup was taken may be required. Therefore backup of the UNDO tablespace should be taken as well (see the explanation for this assumption below). However browsing the internet (including the Oracle online documentation) I did not find any statements concerning the backup of the UNDO tablespace when an online backup is taken. Moreover I couldn't figure out when exactly such a backup of the UNDO tablespace must be done, to ensure that the database can be recovered in all scenarios.
I believe that undo information from the time the hot backup was taken may be required e.g. in the following scenario:
Assume we are taking a hot backup of a given tablespace, i.e. we are copying all OS files that make up this tablespace, while the database is potentially being used. Let D1 be one of the datafiles in our tablespace and let transaction T1 modify datafile D1. Let transaction T1 further be uncommitted while the copy of datafile D1 is being made and let (at least some of) the changes from T1 be included in the backup copy D1' of D1 (because DBWR has already written the modified blocks at the time they were being copied to the backup). Let transaction T1 be rolled back after the copy is completed. D1' will thus contain modifications from T1, while D1 will not.
Now some time later the datafile D1 is lost. When recovering D1 from the copy D1', the (archived) redo logs will be applied to D1'. Before that, transaction T1 should be rolled back in the copy D1', because modifications from T1 must not appear in the recovered version of the database.
I do however not understand, where the information to rollback transaction T1 exactly comes from. It may still be in the current UNDO tablespace. I do however assume that rollback information is not kept in the UNDO tablespace forever. I see three possible answers to this
(a) There are some requirements which I missed so far to backup the UNDO tablespace whenever a hot backup is made.
(b) Since the Oracle "RDBMS" has to be notified that an online backup is being done, it might store all relevant undo information (e.g. write it to the redo log) when the tablespace is put in backup mode.
(c) There are situations when recovery is not possible due to "missing old UNDO information".
Answer (b) seems the most plausible to me. I did however not find any confirmation of this and if (b) really is the answer, I would be interested to understand what information is stored where by the Oracle RDMBS and how it is used for recovery.
To summarize I have the following questions:
(I) Is there any requirement to backup the UNDO tablespace together with an online backup of a tablespace, and if so, where is this stated in the Oracle documentation?
(II) What mechanisms ensure that uncommitted transactions can be cleared from the online copy of a tablespace (potentially a long time after the copy was taken)?
(III) Do you know any links (Oracle documentation or other online resources) explaining these datails?
Thank you for any hints and answers
kind regards
Martin

Its a highly technical question and I can be completely wrong due to my very less knowledge but I would attempt to answer still. Hope I say something sensible so bear with me.
As far as I understand, an online backup of a tablespace can be performed by copying the OS files making up a tablespace while the database is up and being used (i.e. transactions are modifying data in the database).Correct. But it would depend on the tool you are going to use to do so.Using o/s level commands like CP and all would require you to manually copy the files to the backup location. Using RMAN, it would be lot easier and RMAN would take care of everything.
Before the copying of the OS files starts, the Oracle RDMS must be notified that an online backup is being taken via "ALTER...BEGIN BACKUP" (such that some additional information is written to the Redo Log, which may be required for subsequent recovery using the online backup). Again, this is a requirement only in the case of user-managed backup . In that case, because of the fractured block issue , its important that the corresponding older information/image of the buffer is also copied in the redo stream and that's done when the begin backup command is used. Using RMAN, this is not needed as RMAN can read the consistent image which it would store in the backup piece, exactly in the same way in which select request is fulfilled by oracle for a dirty buffer which is yet to be made consistent.
During recovery the Oracle RDBMS uses the copies of the OS files together with the online and archived redo logs in order to reconstruct all committed transactions and it further uses the UNDO tablespace to rollback open (uncommitted) transactions.Correct!
Thinking about this, it seems to me, that in order for this to work in all possible scenarios the undo information from the time the backup was >taken may be required. Therefore backup of the UNDO tablespace should be taken as well (see the explanation for this assumption below). >However browsing the internet (including the Oracle online documentation) I did not find any statements concerning the backup of the UNDO >tablespace when an online backup is taken. Moreover I couldn't figure out when exactly such a backup of the UNDO tablespace must be done, to >ensure that the database can be recovered in all scenarios.The reason that its not a must to do so is that if the transaction is yet active, there is no way that Oracle would overwrite the Undo information of it, even if you may come after 100 years, it would be there. The Undo segment would mark those blocks as active undo blockswhich contains the information of that transaction whose status within the transacton table of that undo segment is still marked as active. So its there all the time in the undo tablespace. Now, for an instance, let's assume that the undo is not there as well( it would be but let's assume), the changes made to the undo segment's blocks are also recored in the redo as its just a change happening to any other segment like EMP,DEPT except with the difference that its not done by you but by oracle. So using that information, in the future , if there would be a need to replay those changes, necessary information to do so can be brought up from the redo blocks stored in the redo/archive logs. Yes, if there would be pending transactions that would require Undo information to get them rolled back and you have lost Undo tablespace and have no backup of it , you wont be able to bring back the database as it would be inconsistent and oracle would not let you to do it. In that case, you may require to use hacks to get it up and that would be really tricky situation.
(I) Is there any requirement to backup the UNDO tablespace together with an online backup of a tablespace, and if so, where is this stated in the Oracle documentation?As I said above, it must be there if you are anticipating loss of Undo tablespace. If you have lost it, you would need a backup and all the archive logs and redo logs to recover and get it back to the point where the current database is . Rest, oracle would take care as it would reapply the redo contents of the undo segments over the undo segment and get it consitent.
(II) What mechanisms ensure that uncommitted transactions can be cleared from the online copy of a tablespace (potentially a long time after the copy was taken)?As I said , pending transaction's undo is never overwritten by Oracle. Its always kept and marked as active undo . Only a transaction end would make it elgible to get overwritten and that too won't happen immediately(undo_retention would kick in) .
(III) Do you know any links (Oracle documentation or other online resources) explaining these datails?I have to see if its there some where step by step mentioned and I shall update the reply once I shall find the link. Hoep someone else in the meantime finds it .
HTH
Aman....

Database open fails after online backup recovery

Hi Friends
We are trying to set up an additional server using the online backup of our DEV server. We have been following SAP Note 549828 for the same. Having restored the online backup, the open database failed.
To resolve the same, in accordance to SAP Note 549828, we created a backup control file with success using the command
create controlfile reuse set database DEV resetlogs noarchivelog
However on issuance of the command
RECOVER DATABASE USING BACKUP CONTROLFILE UNTIL CANCEL
we run into an error as to
ORA-00279: change 794638222 generated at 10/25/2007 12:43:20 needed for thread 1
ORA-00289: suggestion : /oracle/DEV/oraarch/DEVarch1_9766.dbf
ORA-00280: change 794638222 for thread 1 is in sequence #9766
Specify log: {<RET>=suggested | filename | AUTO | CANCEL}
ORA-00308: cannot open archived log '/oracle/DEV/oraarch/DEVarch1_9766.dbf'
ORA-27037: unable to obtain file status
SVR4 Error: 2: No such file or directory
Additional information: 3
ORA-01547: warning: RECOVER succeeded but OPEN RESETLOGS would get error below
ORA-01194: file 1 needs more recovery to be consistent
ORA-01110: data file 1: '/oracle/DEV/sapdata1/system_1/system.data1'
We even manually copied the file system.data1 from the source to the target server but to no avail.
Also the SQL command
SELECT FILE#, CHANGE# FROM V$RECOVER_FILE
displays a different change# for system.data1 while it is showing same number for all other datafiles.
Please advice at the earliest as we are struck. Points awaiting their master.
Regards
Lokesh Gupta

Some inputs addition to Erics comments
The problem is you dont have archives ie Offline redo logfiles in correct location.
/oracle/DEV/oraarch/DEVarch1_9766.dbf Here DEVarch1_9766 is archive file which is misssing from location /oracle/DEV/oraarch. To recover DB you need Archives generated during hot backup.
Generally these steps willl give you desired result.
select * from v$logfile;
We normally switch the log files to the number of log groups that exist
alter system switch logfile;
create a Backup directory to hold our hot back datafiles and Archives
When the backup is complete check the backup location to see if all the files are available. We could now either FTP the same to the other system or copy over these files to another location in case of cloning on the same system.
Copy over all the files to their respective filesystems and directories and then edit the file that was created using the backup controlfile to trace. Copy that file to the remote system and edit it accordingly.
check that all the files are in the right location and edit that information in the control file
Once the controlfile is run successfully and you get the statement processed, we can start applying the archive logs that we have moved to the archive log destination directory as per the init<sid>.ora file.
do a recover of the database to its consistent state
recover database using backup controlfile until cancel;
The create control file command only changes the structure of the database and the SID name, the header of the datafiles still hold all the required information. The above command would ask you to input the archive log file names one by one to do recovery or you could choose the AUTO option. Once the recovery process is complete, open the database with the resetlogs option
Regards
Vinod

Problem in new database creation with the help of online backup

Dear dba's
i am using oracle 11gR2 database in windows server 2003. database is running in ARCHIVE LOG mode.
i have taken an online backup of all datafile,controlfile and spfile.Then i crated folders in all the locations as required for new database.
then i registerd the service of new database named as 'newdb' by
oradim -NEW -SID newdb
then i created a password file manually in 'oracle_home\database' location.
i created a new contolfile named as controlfile_01.ctl. the content of controlfile as follows
STARTUP NOMOUNT
CREATE CONTROLFILE SET DATABASE "NEWDB" NORESETLOGS ARCHIVELOG
MAXLOGFILES 16
MAXLOGMEMBERS 3
MAXDATAFILES 100
MAXINSTANCES 8
MAXLOGHISTORY 292
LOGFILE
GROUP 1 (
'D:\APP\ADMINISTRATOR\ORADATA\NEWDB\ONLINELOG\O1_MF_1_7FK0XG7B_.LOG',
'D:\APP\ADMINISTRATOR\FLASH_RECOVERY_AREA\NEWDB\ONLINELOG\O1_MF_1_7FK0XHWB_.LOG'
) SIZE 50M,
GROUP 2 (
'D:\APP\ADMINISTRATOR\ORADATA\NEWDB\ONLINELOG\O1_MF_2_7FK0XKB8_.LOG',
'D:\APP\ADMINISTRATOR\FLASH_RECOVERY_AREA\NEWDB\ONLINELOG\O1_MF_2_7FK0XM0Z_.LOG'
) SIZE 50M,
GROUP 3 (
'D:\APP\ADMINISTRATOR\ORADATA\NEWDB\ONLINELOG\O1_MF_3_7FK0XNOZ_.LOG',
'D:\APP\ADMINISTRATOR\FLASH_RECOVERY_AREA\NEWDB\ONLINELOG\O1_MF_3_7FK0XOWB_.LOG'
) SIZE 50M
DATAFILE
'D:\APP\ADMINISTRATOR\ORADATA\NEWDB\DATAFILE\O1_MF_SYSTEM_7FK0SKN0_.DBF',
'D:\APP\ADMINISTRATOR\ORADATA\NEWDB\DATAFILE\O1_MF_SYSAUX_7FK0SKPG_.DBF',
'D:\APP\ADMINISTRATOR\ORADATA\NEWDB\DATAFILE\O1_MF_UNDOTBS1_7FK0SKTC_.DBF',
'D:\APP\ADMINISTRATOR\ORADATA\NEWDB\DATAFILE\O1_MF_USERS_7FK0SKWB_.DBF',
'D:\APP\ADMINISTRATOR\ORADATA\NEWDB\DATAFILE\O1_MF_EXAMPLE_7FK0Z5LK_.DBF',
'D:\APP\ADMINISTRATOR\ORADATA\NEWDB\MARSH.DBF',
'D:\APP\ADMINISTRATOR\ORADATA\NEWDB\JOMARSH.DBF'
CHARACTER SET AL32UTF8
the control file path was registered in pfile also.
then i brought the database to nomount stage.
the problem is when i try to mount database it shows following error. anyone can help me to over come from this issue????????
SQL> startup pfile='D:\app\Administrator\product\11.1.0\db_1\database\INITnewdb.ora' nomount;
ORACLE instance started.
Total System Global Area 535662592 bytes
Fixed Size 1334380 bytes
Variable Size 301990804 bytes
Database Buffers 226492416 bytes
Redo Buffers 5844992 bytes
SQL> ALTER DATABASE MOUNT;
ALTER DATABASE MOUNT
ERROR at line 1:
ORA-00205: error in identifying control file, check alert log for more info
the alert massage is:
ORA-00210: cannot open the specified control file
ORA-00202: control file: 'D:\APP\ADMINISTRATOR\ORADATA\NEWDB\CONTROLFILE\CONTROLFILE_01.CTL'
ORA-27048: skgfifi: file header information is invalid
OSD-04001: invalid logical block size (OS 1413563730)
Fri Dec 09 13:11:59 2011
Checker run found 1 new persistent data failures
ORA-205 signalled during: ALTER DATABASE MOUNT...
Thanks & Regards,
John Marshal.A

Hi;
Error: ORA 205
Text: error in identifying control file <name>
Cause: The system could not find a control file of the specified name and
size.
Action: Either
Check that the proper control filename is referenced in the
CONTROL_FILES initialization parameter in the initialization parameter
file and try again.
When using mirrored control files, that is, more than one control file
is referenced in the initialization parameter file, remove the control
filename listed in the message from the initialization parameter file
and restart the instance.
If the message does not recur, remove the problem control file from
the initialization parameter file and create another copy of the
control file with a new filename in the initialization parameter file.
Regard
Helios

Online Backup Failed from DB13, brbackup runs on application server

Hi Everone,
I am getting problem when i am scheduling Online Backup from DB13 t-code.
Backup gets failed because logical command BRBACKUP execute on our Application server host rather than DBCI server host.
I have pest the lob logs below:
Job started
Step 001 started (program RSDBAJOB, variant &0000000000389, user ID BASIS)
Execute logical command BRBACKUP On host xyzapp - (Application server)
Parameters:-u / -jid ALLOG20110824210000 -c force -t online -m all -p initSEP.sap -a -c force -p initSEP.sap -s
d
BR0051I BRBACKUP 7.00 (40)
BR0252E Function fopen() failed for '/oracle/client/10x_64/instantclient/dbs/initSEP.sap' at location BrInitSapRead-1
BR0253E errno 2: No such file or directory
BR0159E Error reading BR*Tools profile /oracle/client/10x_64/instantclient/dbs/initSEP.sap
BR0280I BRBACKUP time stamp: 2011-09-02 21.01.09
BR0301E SQL error -12545 at location BrDbConnect-2, SQL statement:
'CONNECT /'
ORA-12545: Connect failed because target host or object does not exist
BR0310E Connect to database instance SEP failed
BR0056I End of database backup: begroqyn.log 2011-09-02 21.01.09
BR0280I BRBACKUP time stamp: 2011-09-02 21.01.09
BR0054I BRBACKUP terminated with errors
External program terminated with exit code 3
BRBACKUP returned error status E
Job finished
Please help.
Thanks,
Ocean

Hello
The previous contributor is correct, the profile is usually read from ORACLE_HOME. I have never seen this error linking to the Oracle client directory, please double check the following
Ensure that your envirionment variables are correctly set ORACLE_HOME, DIR_LIBRARY
Brtools are always installed on the DB server and can be called from any application server. if you are scheduling from DB13 on the application server, what is the connection setup? RSH, RFC, standalone gateway?
If using a standalone gateway, it must be installed on the DB server, please refer to oss notes #446172, #1025707
Best Regards
Rachel

ONLINE Backup from DB13

Hello :
When we try plan online backup and offline backup we obtein next result,
BR0252E Function fopen() failed for '/oracle/CR5/sapbackup/bdxxipoo.log' at location file_printout-4
BR0253E errno 2: No such file or directory
External program terminated with exit code 5
We have sap instance and DB machine in differents machines.
Please could you help us ???
Thanks

Hello,
check the authorization to
/oracle/CR5/sapbackup/

Oracle offline backup in cluster setup

Dear Experts,
I have installed CRM 2007 on windows and oracle 10g with High availability setup (cluster) and currently in the process of Post-Installation.
As part of post installation, I have to apply support patches. Before that I wanted to take one backup of the system and tried taking the offline backup. My SAP and DB resources are on Node A and before starting the backup I have gone through the "Note 378648 - Offline backup support with Oracle Fail Safe"
and have set the environment varaibles on both Node A and Node B.
Have started the backup from DB13 and initially backup started and DB resource went offline and backup started and files are getting backedup in the path given. But immediately all DB resources came online on Node A and was able to login to SAP. Strange part is even after DB was online backup continued to copy the files in menitoned path and at the end backup completed with errors stating that instance was started during the offline backup and could not start already running instance.
Online backup is working fine. Please suggest how to sort out this issue.
Thanks & Regards,
Sharath

Hi,
The good news is that your cluster works perfectly for fail overs !
The bad news is that, there is something wrong in your configuration because the database was not set offline as a cluster ressource. Do you have a standalone gateway on the Oracle node ?
Regards,
Olivier

Less redo generates at the time of online backup using RMAN

Hi,
can someone eleborate that why less redo generates at the time of RMAN online backup in comparison of manual hotbackup.
Thanks

A manual hot backup (aka "OS backup" or "scripted backup") cannot handle fractured blocks. This happens when a backup command is copying out blocks from a datafile at the same time as DBWR is updating blocks. The Oracle Block Size (8KB) is generally a multiple of the OS Block size (512Bytes to 4KB or 8KB, depending on the filesystem / raw device and how the OS command copies blocks out). It is possible for DBWR to write 8KB and the OS backup utility to be concurrently reading the same Oracle block as, say, two different 4KB reads. The first half of the Oracle block may be read by one OS call and the second half by another OS call -- the two halves would be inconsistent with each other if Oracle were writing the database block at the same time.
(Therefore a restore of such a backup would give you an inconsistent -- i.e. fractured -- block, which would be logically corrupt).
In order to handle such issues, after a BEGING BACKUP is issued, for every first change to a database block, the whole block is written to the redo stream, not just the change vectors. (If the same database block is subsequently modified while in BACKUP mode, it is only the change vectors that have to be written to the redo stream). Once the END BACKUP is issued, the redo stream does not need to capture whole database blocks.
If such a backup is restored, it still has fractured blocks. However, when applying the ArchiveLogs generated during the Backup, each such fractured block is "automagically repaired" because a whole image is restored from the ArchiveLog. (Subsequent changes are replayed as changes through the change vectors).
Hemant K Chitale
http://hemantoracledba.blogspot.com

Online Backups Rman

HI,
I have setup an oracle database running one of our inhouse applications on a Linux Red Hat 4 Enterprise OS. Project go live
is a few days away and I don't have time to teach myself Oracle backups at this stage. I was hoping a few experts could
help me out with the following.
1. What configuration settings do I make to Oracle database and RMAN to prepare for online backups? How do I check
that the settings have already been done?
2. what commands do I then run to trigger off a online backup and what script type do you use so it can be scheduled?
3. How do you schedule the online backup?
My plan its to do an online backup to a folder on another partiton and then use rsync to sync the backup over to another
server which has tape backup.
I also though of running a script to shutdown the databse, then use rsync to copy the databse to a new location and then
startup the database again. Anyone ever use this backup type?
Many Thanks!
Raikal.

Hi 581156,
You can use RMAN to do your online backups, please can you tell us your RDBMS version,OS and if the DB is on archive mode or not? I need this to send you the right links to help you on this process.
Please take a look on this link:
http://www.orafaq.com/wiki/Oracle_database_Backup_and_Recovery_FAQ
Also if you need more information, please refer to:
http://tahiti.oracle.com/
Cheers,
Francisco Munoz Alvarez

Oracle Online backup limitations

Similar Messages

Maybe you are looking for