Data Guard :: Failover / Switchover Without Loosing Previous RMAN Backups
Apr 5, 2013
Dataguard and RMAN.
Got Active DataGuard on a primary database, quite nicely sending its archive logs to its secondary. I can quite happily use the Broker and switchover between them.
Now if I take RMAN backups of the primary database, if I have to failover to the secondary, I'm gonna loose all those backups.
Well, I can restore the whole database to the backup, cos I can restore the control file, from the backuip and therefore I can restore the whole db.
But if I want to restore to a tablespace, I wont be able to , cos the db_unique_names names are different, and the DB ID's will be different.
Same goes if I use a recovery catalog....
so how do I failover/switchover without loosing my previous rman backups ??
Here we have a dataguard environment with db1 as primary and db2 as physical standby database. We configured dataguard in 10g environments and we are not using grid control. What are the steps for configuring rman for incremental backups in this environment?
We want to schedule the backups to be taken in standby db, (that is applicable in case of switchover too). In the case of switchover to db2 what all do we need to do in rman manually to resume backups in db1?
In my production -- we have 3 nodes Primary RAC with 2 Nodes Physical secondary Standby RAC. Due to some reason we will have to failover means 3 nodes Primary Standby RAC to 2 Nodes secondary physical Standby RAC. share Steps for Failover ( not switchover) from 3 Nodes Primary RAC to 2 Nodes Physical Standby RAC. I know this is generic and can be find from google but issue is how can failover from 3 nodes primary standby rac to 2 nodes secondary physical standby rac.
I run DLY cold backups including controlfile & archivelogs and keep 3 gens of backups and archivelogs to 3 days also. What is the correct procedure to restore back to an old cold backup. Database is backed up Mon,Tue,Wed on Thur we require to bring database back to Tuesday DLY security.
I have tried: startup mount restore database from tag 'DLYTUE' recover database
( this just rolls forward on all archive logs once it restores datafiles, not getting to point to alter database open resetlogs.)
Next tried to :
recover database archivelog tag 'DLYTUE' ( same outcome )
What is the correct rman procedure including should I remove old archive logs from working OS directory ? Also need to know once I restore back and open database to TUE security how do you tidy up the archive logs and dumps which are no longer require i.e. : weds..
I have read about using the set until time option but how to I get absolute time from when cold backup completed if this is the correct procedure.
- Oracle 11.2.0.3 - RHEL 4.6 64 bits - Active Data Guard - Broker configured - DBconsole configured on primary server
Now, I have server PRIM with the primary database, and server STB with the standby server. For now, the dbconsole is configured in the PRIM server.When I do a switchover (to the standby on STB), EM is not configured, so I use RepManager.sh to clean the sysman user, and I use emca to recreate the dbconsole. After this, everything works.
Once I switch back, the EM is no longer configured on PRIM, so I repeat the previous steps (repManager and emca). Everything is ok after this. I can access the EM on the primary normaly, and this gets information from the broker to provide basic information on the STB (like db status, apply, and so forth).
I have a dataguard configuration (physical standby).Database A is the primary database and database B is the standby database (I do not use the broker).Both are Oracle RDBMS 12cR1 EE. The Apply process uses the current log (standby redo logs).On database B, when querying v$managed_ standby;, I get:
It works fine. Before to perform a switchover,- from database A (current physical database), I get switchover_status = 'TO STANDBY'- from database B (current standby database). I get switchover_status = 'NOT ALLOWED'. What switchover_status = 'NOT ALLOWED' means exactly and how to get further with the switchover?
we have a created primary and physical standby databases. Before the switchover the application connected to dataguard allows updates on the database. However after the switchover, it allows read-only access and we are unable to update. Any comments prima-facie ?
I am fairly new to Oracle and have recently started learning the basics of Data Guard. I have set-up a Primary 11g database which successfully ships and applies logs to a Single Standby (on the same server) as expected. Manual switchovers work fine; however, I have found a slight problem when performing switchovers using the Data Guard Broker utility.
The switchover is successful- but the broker cannot restart the Primary instance. The strange thing is that when I perform a switchover in the opposite direction (i.e. the new Primary goes back to standby again) it works fine. The error is receive is: ORA-12514: TNS:listener does not currently know of service requested in connect descriptor.
The majority of posts I have seen on this topic seem to point to incorrect listener / tnsnames set up but they have exactly the same values. If both databases are identical, why does it work one way but not the other? I will provide any necessary files to pinpoint where the issue could be.
i have a db's in maximum protection mode and i am trying to perform a switchover of physical standby database to primary role. but it failed., the steps followed is as follows. everything is fine with old primary site and the it is now in mounted state. When I went through the log file, I got the error message as "ORA-16072: a minimum of one standby database destination is required".the standby_archive_dest is as follows.
old primary database
SQL> alter system switch logfile; SQL> alter system archive log current;
old standby database...Disconnect the managed recovery mode in Standby database.
SQL> recover managed standby database cancel; Media recovery complete. SQL> recover standby database; AUTO SQL> exit [code]....
I am having setup of 2 node 10.2.0.4.0 RAC setup on ASM with Single Instance Physical standby on File System (OS: RHEL 5.4). I configured DG Broker without any major issues. Getting all output of 'show configuration' and all DG related logs are clear. Archives are getting transferred from both Primary nodes.
Now I am getting error Error: ORA-16535: CRS is preventing execution of a broker operation.
DGMGRL> switchover to livedg; Performing switchover NOW, please wait... Error: ORA-16535: CRS is preventing execution of a broker operation
Oracle Says
Quote:Cause A broker operation was underway that required CRS to stop managing the instances of this database, but CRS management could not be stopped on behalf of the broker"s request on at least some of the instances, so the broker operation was canceled.Suspend CRS management of this database using SRVCTL STOP DATABASE -D <name> -O NONE. Then reissue the broker request.But this command is not working as it's given to stop the database.
One more thing, it's nowhere said to stop CRS while performing Switchover / Switchback using DG Broker. When we do Switchover / Switchback we keep open only one of the Primary instance.
is there any documentation / link available to switchover of RAC database using dgmgrl (DG Broker)
I am very new to data guard...I have a primary and a physical standby database...If the primary got crashed, are the users directly switched to physical standyby (is it automated) or the DBA has to manually do the fail over or switch over....is it the same concept as in RAC like TAF....How can the DBA know that the users are disconnected from the database.....explain with some steps...
I am running some small tests here in my test env, using dataguard. I have configured the Primary and Standby with Maximum Availability... they're running just fine. Now i want to execute a failover test (i have already ran a Switchover test with the Broker successfully).
My question is very simple, at my point of view: What are the required steps to execute a successful manual failover? For example, i have my env as follows:
- Primary: prim1 - Standby: stdb1
Suppose that the primary database crashes in an unrecoverable way... is this case a manual failover would be necessary.
To do so, i would have to execute the following command, in my Standby database:
-- stdb1 is the standby database... DGMGRL> failover to stdb1
The above command is correct? Are there any required configurations after the failover? I read the Oracle Docs, and it says
what is the configuration dba has do for enabling the "fast-start failover concept" in the event of loss of primary database without requiring any manual steps to invoke the failover.whether the configuration are to be made in parameter file if yes what are those ?
My Data Guard 's Enviroment: Primary: 11.1.0.7 RAC 2 Nodes Standby: 11.1.0.7 Single Instance
I want to implement Fast-Start Failover at my enviroment. Reading requirements, I found that Oracle Flashback is needed (I don't use Flashback on my databases).
Best Practices recommends 60 min (1 hr) for DB_FLASHBACK_RETENTION_TARGET if I just want to use Flashback to fast-start failover.
My questions:
What 's better, setup up flash recovery area on ASM or O.S. filesystem? What is the recommended size for flash recovery area?
I have a situation where I want to configure primary database (11.2) with 2 remote destinations. dest_2 is the default and points to a standby on host_2. However, I also want the primary db to continue transporting redo to dest_3 on node_3 when node_2 is taken down (planned or unplanned).
1) Configure the ALTERNATE attribute of dest_2 to point to dest_3.
2) Configure tnsnames client-side failover on primary host to point to 2 nodes (node_2 and node_3).
I have set up a single instance standbys for rac databases. Now, I need to offload the backups from the primary on to the standbys.
1. cancel managed recovery
2. connect to target(standby) and catalog and backup the standby
3. put standby in managed recovery mode.
I think following these steps, I can restore primary(?). Now, My question is, how can I use these standby backups to clone/refresh databases? I tried it by connecting to target(primary), catalog and auxiliary but rman is using the primary backups instead of standby's to refresh the auxiliary database.
i have Oracle 10g data guard set up on windows environment.....i need to know the what are the right steps to perform failover in case the primary database gets fail.
If flashback is enable in physical standby database 1. If we failover at 11AM can I flash back NEW primary database to 6 AM ? 2. if I convert physically standby database to snapshot standby database at 11AM , Can I flashback snapshot standby database to 6 AM and do some works on it (DML operations) then converting the snapshot standby database into physical standby database ?
i work in an application that should make the replication from a publusher table to a remote subscribe table, using snapshot,and trigger, replication data of update works perfectly (update,insert,delete), but when i try to add or dropp a clumn in the publisher table, repplication fail, i know that my method d'ont replicate ddl statment like create or alter table, so i would like the better way to do the replication of the ddl statment without loosing tha data in the subscribe table, i'm working with oracle XE,
"If no level 0 backup is available, then the behavior depends upon the compatibility mode setting. If compatibility is >=10.0.0, RMAN copies all blocks changed since the file was created, and stores the results as a level 1 backup. In other words, the SCN at the time the incremental backup is taken is the file creation SCN. If compatibility <10.0.0, RMAN generates a level 0 backup of the file contents at the time of the backup, to be consistent with the behavior in previous releases."
So yes, that's it. Even in the Oracle 11g OCP course and exam, the information given is for Oracle 9 (pre 10g). Seems like a major functional regression. ------------------------------------------------------------------------------------------------------------------------------------------------ Original problem.
Am I not understanding something about rman???? Using oracle 10g standard edition and rman.
No existing backups.
When I run repeated level 1 cumulative incrementals, the appear to back up everything (like a level 0 would). My understanding is that if a level 1 is run w/o an existing level 0 backup, it will generate a level 0 backup. All subsequent level 1 backups should be level 1's as expected.
If I explicitly generate an level 0, followed by level 1's it all works as expected.
I am determining what got backed up by the size of the resulting save sets.
Do I have to do an explicit level 0 and then explicit level 1's? I thought not.
if I run below rman command lines in 10g and 11g RMAN, it works well, but NOT in RMAN (Release 8.1.7.4.0), so which commandlines I should use to get to the same pursposes in RMAN 8.1.7.4?
run { ALLOCATE CHANNEL c1 TYPE 'sbt_tape'; ALLOCATE CHANNEL c2 TYPE 'sbt_tape'; ALLOCATE CHANNEL c3 TYPE 'sbt_tape';
I am trying to create the Standby database using RMAN.
I have first created an auxiliary instance followed by the password file for the instance and the registering it it listener.
Below are the parameters of the pfile of the standby database and followed by the command issued at the RMAN prompt.
PFILE parameters [STANDBY] orcl.__db_cache_size=293601280 orcl.__java_pool_size=8388608 orcl.__large_pool_size=8388608 orcl.__oracle_base='F:appdsharma'#ORACLE_BASE set from environment
Can i setup physical standby without RMAN? I want to create a fresh single physical standby without RMAN .. If i copy all files in another system and take the controlfile from primary as alter database create controlfile for standby ... and fill all the gaps with archives.. Should it work ? i am on 10G Rhel 4 linux.. and both machines would have same config
We have two Windows 2008 VMs with Oracle 11G configured as primary and standby.This configuration has been defined by a partner who is no longer working for us.We have broken the correct application of the redo log on the standby when we have not correctly restored some snapshots on those two VMs.I would like to create a new Oracle server that will become the new standby. For that, I need too :
- Display the full dataguard configuration to reproduce it on the new server - Stop (delete ?) the configuration on the primary server.
We also have the current problem that RMAN refuses to delete old archivelog since they have not been applied to the standby.
