DataGuard configuration

Similar Questions

• demote the Group patch 3 10.2.0.5 in dataguard configuration

  Hello

  We run in running in linux x86_64 5 update 7 dataguard configuration.

  primary database "single instance": 10.2.4.4 (downgraded to 10.2.0.5)

  STANDBY database "2xnode CARS": 10.2.0.5 (need to downgrade)

  primary database successfully downgraded after slaughter again shipping services, etc...

  in 10.2.0.5 readme group of hotfixes, I can't find a note dealing with the ' how to make to delete/uninstall only from binaries group of patches as in dataguard we must not to run one of the scripts what database keys.

  thanking you in anticipation.

  Kind regards

  -Ahmad

  Hi Andre,

  It is not possible to uninstall a group of patches. Either Oracle home must be restored from Oracle home or install it again. Once this is done, you can simply enable log shipping and it should take care of database backup.

  334598.1 - how to restore a group of patches

  Thank you

  Ashish

• Dataguard Configuration database upgrade

  Hello
  
  I have some doubts regarding the upgrade of database in dataguard configuration. Could someone help me please.
  
  -How update occurs in the dataguard configuration?
  -Is that we must install only different OH in both? And once the catupgrade script is run on the primary, it will automatically update pending as well?
  - Or is to upgrade the two separately? That is to say, first aid and then primary or vice versa
  -Can reduce us downtime by breaking the dataguard config. Do a permutation, and do secondary as main and primary upgrade first?
  And once the upgrade is complete in the primary, raise and improve the day before. And later once the upgrade completed on standby, primary synch and sleep.
  
  Please share your ideas and your thoughts!
  
  Thank you!

  First of all, - steps may differ about the version upgrade but - my post with the following link might help clarify the concept.
  http://emrebaransel.blogspot.com/2009/02/10204-upgrade-with-DataGuard.html

  user1982050 wrote:
  Hello

  I have some doubts regarding the upgrade of database in dataguard configuration. Could someone help me please.

  -How update occurs in the dataguard configuration?

  First you patch the binary Oracle both primary and standby. Then open services dataguard and upgrade the primary, standby database will update already.

• DataGuard configuration in production

  Hi friends,

  I have to configure my production environment (primary) to a database of pending. I have a question for this I need to make some changes in the file(spfile/pfile) parameter any time will I have to bounce my primary database or it need the downtime?

  Thank you and best regards,

  Arya

  Hello

  To configure dataguard, must actually DB bounce we configure dataguard in my company without rebound. All the required parameters can be defined using the alter system command to set the primary database memory, even if you use on startup pfile. When using pfile don't forget to edit to carry all the settings that have been defined on the primary database to configure dataguard so that when the DB is rebound later or to scheduled maintenance window, the settings will be retained and it posed problems for the installation of dataguard.

• Agent upgrade on the DataGuard configuration

  Hello
  
  Win x 64 server
  Oracle 10.2.0.3.0
  Agent 10.2.0.3.0 (go 10.2.0.5)
  
  Last days I've been patching all the agents out there, nothing complicated on this matter, but now I came to our configuration dataguard, for me, it seems that this will not matter, I have found no documentation contends it would be different, since the agent are separated from data bases itself.
  
  #shut to the bottom of the agent
  #patch
  
  Grateful for answers,.

  There is nothing special dataguard / agency relationship. As usual, upgrade after upgrade test, and if you find any problems you will need to remove the target, and add it again.

• RMAN archive deletion problem in the dataguard configuration

  I have a weird problem in all our databases with dataguard. When we issue "delete archivelog all saved on the drive; 1 times"the archiving that have not been backed up logs are also appear as candidates for deletion.

  To confirm these observations, I did a test in our lab.

  (A)

  On a stand alone DB, I took a backup and then move the log files.

  archivelog all list, any backup archivelog 1 times the disk device type from the list and delete archivelog all saved 1 time on device type disk - all these three orders the results expected. Especially the third command showed only archived newspapers backed up as candidates for deletion

  (b) then I installed dataguard, save on primary and switched some log files.

  Now, list archivelog all and list archivelog all saved 1 time on device type disk yielded expected results.

  BUT delete archivelog all saved 1 time on device type disk shows the newspaper archive, even those that have not been backed up as candidates for deletion. It's curious me as far as command list gives the current results.

  
  

  Everyone saw this behavior. Let me know if you need any o/p for this problem. Thanks for your help...

  If someone is after that, there is a bug/patch for 11.2.0.4. Patch has solved the problem for me.

  

  Bug 16692232 - RMAN backup archivelog all not backed up / remove any entry does not work as expect you (Doc ID 16692232.8)

  Down

• How to disable the dataguard configuration?

  Hello

  
  

  env: Oracle 10 g 2 (10.2.0.5), RHEL 5.8

  
  

  I created a database waiting for a prim. Now I have to spend the day before the "Prim", stop the original role "prim" and disable all DG related configuration.

  
  

  I think to take following steps (in the order listed):

  
  

  Switch to standby mode (stdby will become prim and prim will become Stdby). Then:
  

  On Stdby (which now is prim):

  Change the setting: Enable, Disable LOG_ARCHIVE_DEST_STATE_2

  
  

  'Set null' all these parameters:

  LOG_ARCHIVE_DEST_STATE_2

  LOG_ARCHIVE_DEST_2

  FAL_SERVER

  FAL_CLIENT

  LOG_ARCHIVE_CONFIG

  The Prim (which isn't Stdby):

  Shutdown

  
  

  It is that OR is it more than that?

  
  

  Please let me know if there is no error or if I missed something?

  
  

  Best regards

  Hello;

  It should work. You need not to stop the primary database for this.

  On Oracle 11, you can use:

  ALTER DATABASE GUARD NONE;

  I don't know if it works on Oracle 10 and all my databases Oracle 10 disappeared.

  Best regards

  mseberg

• ANALYZING listener 11 GR 2 and AMT DataGuard configuration for RAC nodes

  By oracle paper MAA for 11 GR 1 material to configure Oracle Networking, they recommend to configure listener_hostname on each host (2 Guest CARS on primary, 2 Guest CARS standby in standby mode). Like Listener_europa, Listener_callisto, Listener_dione, Listener_hyperion.
  
  I installed two primary databases grid/oracle RAC and now try to configure both nodes to backup CARS. I couldn't find any documents on how to configure the network files. 3 SCAN listeners were now configured and running on the main nodes of the rac.
  
  so the configuration of listner.ora, how to set up listner_hostname? or should I use scan listener?
  
  Can someone tell me exactly how I can configure tns entries in such situation?
  
  Thank you

  Hello;

  So I think Yes 10 5 (per node) If you have two.

  See the document Oracle B28282-02

  2.6.4.6 use of standby Redo Logs and configure the appropriate size

  Best regards

  mseberg

• DataGuard Configuration problem / ORA-16047

  So last night I decided to set up a test of physical database paused. I had everything working properly and when I started to play with the Data Guard Broker I started having a few problems. Now I can't get the logs to be shipped from the primary to the day before.
  
  Version: Primary and Standby
  

  BANNER
  --------------------------------------------------------------------------------
  Oracle Database 11g Enterprise Edition Release 11.2.0.1.0 - 64bit Production
  PL/SQL Release 11.2.0.1.0 - Production
  CORE    11.2.0.1.0    Production
  TNS for Linux: Version 11.2.0.1.0 - Production
  NLSRTL Version 11.2.0.1.0 - Production

  OS: Primary and Standby
  

  [oracle@dgdb0 trace]$ uname -a
  Linux dgdb0.localdomain 2.6.32-100.28.5.el6.x86_64 #1 SMP Wed Feb 2 18:40:23 EST 2011 x86_64 x86_64 x86_64 GNU/Linux

  First of all, I noticed a problem with a gap in the sequence numbers.
  
  Standby
  

  SQL> SELECT sequence#, applied from v$archived_log order by sequence#;
  
   SEQUENCE# APPLIED
  ---------- ---------
       8 YES
       9 YES
      10 YES
      11 YES
      12 YES
      13 YES
      14 YES
  
  7 rows selected.

  Primary
  

  SQL> archive log list;
  Database log mode           Archive Mode
  Automatic archival           Enabled
  Archive destination           USE_DB_RECOVERY_FILE_DEST
  Oldest online log sequence     37
  Next log sequence to archive   39
  Current log sequence           39

  Here are the configuration on the primary information:
  

  SQL> show parameter db_name
  
  NAME                     TYPE     VALUE
  ------------------------------------ ----------- ------------------------------
  db_name                  string     dgdb0
  SQL> show parameter db_unique_name
  
  NAME                     TYPE     VALUE
  ------------------------------------ ----------- ------------------------------
  db_unique_name                 string     dgdb0
  SQL> show parameter log_archive_config
  
  NAME                     TYPE     VALUE
  ------------------------------------ ----------- ------------------------------
  log_archive_config             string     dg_config=(dgdb0,dgdb1)
  SQL> show parameter log_archive_dest_2
  
  NAME                     TYPE     VALUE
  ------------------------------------ ----------- ------------------------------
  log_archive_dest_2             string     service=dgdb1 async valid_for=
                           (online_logfile,primary_role)
                           db_unique_name=dgdb1

  Sleep settings
  

  SQL> show parameter db_name
  
  NAME                     TYPE     VALUE
  ------------------------------------ ----------- ------------------------------
  db_name                  string     dgdb0
  SQL> show parameter db_unique_name
  
  NAME                     TYPE     VALUE
  ------------------------------------ ----------- ------------------------------
  db_unique_name                 string     dgdb1

  So I proceeded to run this query:
  

  SQL> SELECT error from v$archive_dest WHERE dest_name='LOG_ARCHIVE_DEST_2';
  
  ERROR
  -----------------------------------------------------------------
  ORA-16047: DGID mismatch between destination setting and target
  database

  The error description is:
  

  Cause:       The DB_UNIQUE_NAME specified for the destination does not match the DB_UNIQUE_NAME at the destination.
  Action:     Make sure the DB_UNIQUE_NAME specified in the LOG_ARCHIVE_DEST_n parameter defined for the destination matches the DB_UNIQUE_NAME parameter defined at the destination.

  As you can see above the DB_UNIQUE_NAME in the parameter LOG_ARCHIVE_DEST_2 to match that of the database pending.
  
  Also DG_BROKER_START is false on the back-end databases and relief.
  
  Finally, I deleted all the files DRC * the directory $ORACLE_HOME/dbs on primary servers and intelligence to make sure that the broker is not configured.
  
  Where am I wrong? How can I get the standby caught up and working properly again?
  
  I apologized if I missed something. I'm relatively new to databases in waiting.

  Centinul;

  I noticed a few things

  1. If you run the query below the sleep mode, you will probably always get the results you've posted

  SELECT sequence #, applied to the order of v$ archived_log by sequence #.

  What to do if run it from primary school and I add the "DEST_ID" column to the query.

  2. you may have more luck finding GAPS using these queries:

  Select max(sequence#) from v$ archived_log in case of application = 'YES ';

  Choose the process, status from v$ managed_standby;

  SELECT * FROM V$ ARCHIVE_GAP;

  3. you mix SQL result with data broker, which can bite you. You don't know where you went wrong, but I would create PFILE file versions before trying to broker data at both ends. You can view each setting and avoid problems before adding the data broker. Data broker will take control and you can even find that it adds entries to your settings file.

  The ORA-16047 is probably related database parameter and that should help at least answer the question. For example you might miss log_archive_config on the standby or soething. In comparing the two PFILE should refine this

  I checked my notes from data broker, but did not find an ORA-16047, ORA-01031, ORA-16675, ORA-12514 and ORA-16608, I managed.

  For me I decided it was a good idea to run Data Guard without data broker at first, until I got the impression using SQL.

  Finally if you have not already consider the purchase of Larry 'Oracle Data Guard 11g Handbook' Carpenter IMHO its worth every penny and more.

  Best regards

  mseberg

• DataGuard configuration question.

  I need suggestion about the configuration of the dg.
  primary-
  OS - linux
  DB - 10.2
  data files - OMF (db_create_file_dest and db_create_online_log_dest_n)
  FRA is configured.
  archiving of FRA.
  
  Backup site.
  OS - Linux
  DB - 10.2
  already have a stand-alone single instance running on ASM. with two DATA and FRA
  
  So my question is can I create instance watch on ASM?
  primary means no EAM with omf and Eve on ASM for?
  Please notify. Thanks in advance.

  You can... but I wouldn't... and here's why. Assuming that, even if you did not, it is not logical and physical.

  The point of the physical standby, is that it is 100% identical to your production by all means server. Shell scripts, run on one should run without modification or other tests on the other. It is the same for all aspects of infrastructure. Make it different and you invite problems or need to up on all aspects of the tests.

  In case you, to all the less, your upgrade paths, and other names must be different. Of course, I would move one spfile to another.

  Why not move the primary to ASM, and then to implement the strategy of DR?

• EBS Dataguard

  Hi all

  EBS 11i, 9i database

  If I have the dataguard configuration/installation/installation program on our database of PROD, does stop?

  We have already activated archivelog on.

  Thank you very much

  JC

  Yes, you will have to stop

  Follow below

  https://Oracle-base.com/articles/9i/data-guard

  concerning

  Pravin

• DataGuard passage 2

  Hi all

  9i

  RHEL5

  I am show passage on our dataguard configuration.

  The following command are reversed:

  SQL > archive logs list;

  Database log mode Archive Mode

  Active AutoArchive

  Archive destination /u02/oradev/devdb/9.2.0/dbs/arch

  More old sequence of journal on line 113

  Log sequence archive 115

  Current log sequence 115

  SQL > ALTER DATABASE COMMIT to SWITCH mode STANDBY;

  ALTER DATABASE COMMIT TO THE PASSAGE IN STANDBY MODE

  *

  ERROR on line 1:

  ORA-16014: log 3 sequence # 115 not archived, not available destinations

  ORA-00312: wire 3 1 online journal: ' / u02/oradev/devdata/log03a.dbf'

  ORA-00312: wire 3 1 online journal: ' / u02/oradev/devdata/log03b.dbf'

  ORA-16014: log 3 sequence # 115 not archived, not available destinations

  ORA-00312: wire 3 1 online journal: ' / u02/oradev/devdata/log03a.dbf'

  ORA-00312: wire 3 1 online journal: ' / u02/oradev/devdata/log03b.dbf'

  Why is the error?

  I tested before the digital switchover and it was good.

  Help, please.

  Thank you very much

  JC

  Jenna, run swithover using this guide:

  -1 in the original primary:

  change the database for validation to the passage in standby mode with the end of the session;

  stop immediately;

  startup nomount;

  ALTER database base_de_donnees eve of Mount;

  ALTER system set aq_tm_processes is 0 scope = both;.

  ALTER system set log_archive_dest_state_2 = reporter;

  Select switchover_status in the database of v$.

  2 original to standby mode:

  change the database for validation to the passage at the elementary level with the end of the session;

  stop immediately;

  startup

  ALTER system set aq_tm_processes is 5 scope = both;.

  ALTER system set log_archive_dest_state_2 = enable;

  NOW IT IS ELEMENTARY, BUT YOU STILL HAVE NOT ALL EXPECTATIONS!

  -3 primary and future old standby:

  recover managed standby database disconnect;

  check V$ Archived_log and v$ Archive_Gap

  If you find an error, please report it.

  Concerning

• EBS 11i Dataguard Setup

  Hi all

  EBS 11i

  9i

  RHEL5

  There is a special process for the physical database of EBS to standby dataguard configuration?

  I was able to configure the physical standby for its database, but if you try to shift from the primary to the standby mode, the standby cannot log command.

  Help, please...

  Thank you very much

  JC

  Hi all

  I solved my problem now. I'm just being dumb of me

  In the init.ora, I see that there comment #.

  #log_archive_start = true

  Since it's the eve before db, it did not matter if its commented out.

  But when I switch to it, now, the question to which it can not automatically archive the online redo, it is why ALTER SYSTEM SWITCH LOGFILE wait forever

  Thank you ALL.

• Log_archive_dest_2 DataGuard

  Hi all

  9i

  RHEL5

  In the dataguard configuration init.ora parameter, which is the best between the two? In terms of performance?

  LOG_ARCHIVE_DEST_2 = "SERVICE = STBY LGWR ASYNC NOAFFIRM'"

  or

  LOG_ARCHIVE_DEST_2 = "SERVICE = STBY LGWR SYNC STATE '"

  
  

  Thank you very much

  JC
  

  
  

  Hello, if your storage pending system is not the best of all or your /bandwidth of network latency is not very good, then choose ASYNC (and 9i, you can choose ARCH ASYNC... for a primary performance).

  my personal experience:

  300 km of primary education in the various networks... before I used ASYNC.

  Primary and standby on the same data center with a low cost for standby disk storage: I used ASYNC.

  Primary and standby on the same network (with good latency and bandwidth) and the upper disc AND VERY CRITICAL of DATABASE storage!, so I use SYNC.

  With SYNCHRONIZATION option, all of the problems in writing to the backup site are in the PRIMARY.

  Kind regards.

• DataGuard folder names

  Hi all

  9i

  RHEL5

  I am currently dataguard configuration.

  I just copied all our production configuration to the backup target server so everything is identical with its folder names. Their init.ora is who is different.

  The instance_name source is PROD, while the eve instance_name target is STBY.

  My question is: is it OK to remember all the names of folders as is? for example, the folder

  
  

  / U01/oradata/prod where all dbf and control files are located, is it OK to keep such what? one does not change for/u01/oradata/STBY?

  
  

  
  

  Thank you all,

  JC
  

  Hello Jenna:

  Controlfiles are defined by the parameter control_files in initSTBY.ora.

  For the data files (and logs), if db_file_name_convert (log_file_name_convert) in initSTBY.ora. parameter is empty, then PROD and STBY directories must be the same on both computers.

  in the case of initSTBY.ora. you define db_file_name_convert = ' / u01/oradata/PROD ', ' / u01/oradata/STBY ", then all data files who prod/u01/oradata/PROD, in STBY machine are in/u01/oradata/STBY.

  Don t forget to do the same in initPROD.ora, IE.

  db_file_name_convert = ' / u01/oradata/STBY ", ' / u01/oradata/PROD.

  and the same for log_file_name_convert.

  Kind regards.