Eve of DataGuard physical

Similar Questions

• RAC & DataGuard (physical pending)

  Hi all
  
  I am trying to get a high level overview of how CARS & DataGuard behave in the following configuration. I wrote my understanding of how things work. Please correct me if I'm wrong.
  
  (1) node RAC (primary) 2 with a physical standby of single instance.
  
  (a) same init.ora parameters related Eve would have configured on both nodes main cars.
  (b) redo apply service to sleep could merge the redo leave the 2 wires of the primary and apply it to the night before to preserve the synchronization.
  (c) during RAC primary instance must be to the top outside of the instance of Eve.
  (d) during the downgrade as a primary RAC instance should be in place outside the instance of Eve.
  (e) during the failover ofcourse the two primary instances would be down that guarantees the fail on.
  
  (2) node 2 (primary) CARS with a physical standby 2 node
  
  This is when it becomes complex and I'm not really sure how a, b, c, d, e would be in this scenario. Appreciate if you could throw some light on this and the list of what
  a, b, c, d, e could look like in this scenario.
  
  I'm assuming that only one instance in the standby THAT CARS should be in place when the watch is in configuration CARS to do it again apply to the work.
  
  
  Also, if there is a white paper procedure step by step details to put in place the 2 scenarios above, please let me know. So far, I could not find that the MAA white paper, but which was not very helpful. If you prescribe a good book for RAC & DataGuard would be too great...
  
  
  Thanks for your help

  >

  (1) node RAC (primary) 2 with a physical standby of single instance.

  (a) same init.ora parameters related Eve would have configured on both nodes main cars.

  Usually rac nodes share their spfile on a shared diskvolume or through the ASM. So they were all identically DG parameter

  (b) redo apply service to sleep could merge the redo leave the 2 wires of the primary and apply it to the night before to preserve the synchronization.

  Fix

  (c) during RAC primary instance must be to the top outside of the instance of Eve.

  Seems logical to me
  Edit: actually during normal operation the two RAC nodes are up, so during the transition to digital they are both active too
  This is the configuration that is important. As soon as the wait becomes primary, it could be that he only talks to one RAC node for its files of log-apply

  (d) during the downgrade as a primary RAC instance should be in place outside the instance of Eve.

  This single is the only race in step c.

  (e) during the failover ofcourse the two primary instances would be down that guarantees the fail on.

  Fix. Remember that during a failover of the RAC config would be déconfigurée the setting of the DG and must be added again as before.

  You may need to take a look at: http://www.oracle.com/technology/deploy/availability/pdf/MAA_WP_10g_RACPrimarySingleInstancePhysicalStandby.pdf

  >

  (2) node 2 (primary) CARS with a physical standby 2 node

  This is when it becomes complex and I'm not really sure how a, b, c, d, e would be in this scenario. Appreciate if you could throw some light on this and the list of what
  a, b, c, d, e could look like in this scenario.

  I'm assuming that only one instance in the standby THAT CARS should be in place when the watch is in configuration CARS to do it again apply to the work.

  N ° day before primary RAC RAC would mean that both nodes are in place on both sides. PrimNode1 send its logs of StandbyNode1 recovery and the same for PrimNode2 and StandbyNode2

  Have a look here: http://www.oracle.com/technology/deploy/availability/pdf/MAA_WP_10g_RACPrimaryRACPhysicalStandby.pdf

  HTH,

  FJFranken
  My Blog: http://managingoracle.blogspot.com

  Published by: fjfranken on July 16, 2010 01:13

• EBS R12 dataguard physical standby

  Hi all

  BSE R12.2.4

  RHEL 6

  Most of the physical standby of companies used to be opened read only for reports.

  It comes to unload part of the load of PROD for better performance.

  Most of the reporting tools are 3rd party like Jasper, Crystal, Java, .net, Apex, Toad, Sqldev, etc.

  My question is can a simultaneous Manager EBS reports can execute on a database read-only?

  Thank you very much

  JC

  Also

  Using Active Data Guard and discoverer of generation of Oracle E-Business Suite release 12.1 reports and a database in Oracle 11 g or 12 c pending physical (Doc ID 1070491.1)

  Using Active Data Guard of generation of Oracle E-Business Suite 12.2 version reports and databases Oracle 11 g (Doc ID 1944539.1)

• 10g dataguard - physical intelligence database

  Hi guys,.
  
  Db1 < primary - > Production
  DB2 < knows standby - > DR
  
  I need to start my DB2 in DR segment for testing.
  If I make below, will this affect my Production DB? I need it to be in place and works as well.
  
  
  DR:
  recover managed standby database cancel;
  stop immediately;
  startup nomount;
  ALTER database base_de_donnees eve of Mount;
  recover database until cancel; pending
  ALTER database activate standby database;
  
  Thank you

  dbaing wrote:

  My question is, will it decrease my production db if I activate the database pending?

  Thank you

  No, it won't lower db output by activating the standby mode

  It may be useful

  See you soon

• need advice on another Eve in another physical location

  We have a 11.2.0.3 primary RAC two nodes on a location and CARS waiting on another location.
  
  Now, we decided to add another pending in another place, stand-alone node.
  
  How do I set this watch?
  
  Say that our primary db is called BRDF, and then watch the CARS db name is BRDF, db_unique_name is called PRDBS.
  
  WE would like this new db standby on stand-alone server called db as BRDF and db_unique_name is also BRDF.
  
  How to proceed from there if we would use dataguard broker?
  
  Thank you in advance.

  Of course, you only watch, then it can work

  db_unique_name parameters should be different so that the broker and the custody of the data understand and identify the data bases.

• great import on dataguard (physical standby 11.2)

  Hi experts,
  
  I have to import a large database (50 GB) (impdp) which must be configured with a database of pending.
  
  My Question:
  

  Should  I make the import before or after the configuration of the standby database?

  Thanks and greetings
  
  hqt200475
  
  Published by: hqt200475 on 7 Sep 2011 08:55

  Do the IMPORT in the first place.

  If you don't you risk ruining the FRA on the primary and the side of the watch. Roll forward stops applying the second FRA hit the limit of the day before or worse another day or two more later, when you think you're in the clear. Even if you do not use any shipping FRA for a reason this remake is not a good idea on a new system of Data Guard. You might have another question and you just made it by performing an import of this size. On most systems you would add additional 500 more redo logs to the mix.

  Also, you can duplicate using RMAN standby database after import very easily.

  I probably expected a few days after I imported before primary futures a less that I built the day before.

  I have a solid duplicate document and will share if you have decided to go this route.

  Best regards

  mseberg

• addition of relief in dataguard broker database fails

  When adding of Eve in dataguard broker database, I get the following error

  DGMGRL > add database as uat login is uat maintained as physical;

  Error: ORA-16698: parameter LOG_ARCHIVE_DEST_n defined for the object to add

  Failed.

  DGMGRL > output

  I deleted the log_archive_dest_2 parameter in the primary database and null value

  SQL > sho parameter log_archive_dest_1;

  VALUE OF TYPE NAME

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

  Log_archive_dest_1 string LOCATION = USE_DB_RECOVERY_FILE_

  DEST = VALID_FOR (ALL_LOGFILES, A).

  LL_ROLES) DB_UNIQUE_NAME = prod

  SQL > sho parameter log_archive_dest_2;

  VALUE OF TYPE NAME

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

  LOG_ARCHIVE_DEST_2 string

  log_archive_dest_20 string

  Can someone help me please

  vmkumar wrote:

  When adding of Eve in dataguard broker database, I get the following error

  DGMGRL > add database as uat login is uat maintained as physical;

  Error: ORA-16698: parameter LOG_ARCHIVE_DEST_n defined for the object to add

  Failed.

  DGMGRL > output

  I deleted the log_archive_dest_2 parameter in the primary database and null value

  SQL > sho parameter log_archive_dest_1;

  VALUE OF TYPE NAME

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

  Log_archive_dest_1 string LOCATION = USE_DB_RECOVERY_FILE_

  DEST = VALID_FOR (ALL_LOGFILES, A).

  LL_ROLES) DB_UNIQUE_NAME = prod

  SQL > sho parameter log_archive_dest_2;

  VALUE OF TYPE NAME

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

  LOG_ARCHIVE_DEST_2 string

  log_archive_dest_20 string

  Can someone help me please

  What you have shown here is the log_archive_dest_2 of the primary database. You must clear the log_archive_dest_2 of the standby database too.

  The error is speaking on the parameter log_archive_dest_2 of the database pending. 12 c, then the broker configuration, you must clear the log_archive_dest_2 (destinations archive remote) on the primary and standby databases. DG broker would take care of her.

  On autonomy in standby:

  SQL > alter system set log_archive_dest_2 = ";

  And then try to add the standby database to the broker configuration.

  HTH

  -Jonathan Rolland

• DataGuard RAC Setup

  With the help of 4 knots 10.2.0.4 RAC.
  I'm in the process of establishing a node 4 10.2.0.4 RAC dataguard physical standby site. I followed to set up document:
  [MAA RAC physical Standby | http://www.oracle.com/technology/deploy/availability/pdf/MAA_WP_10g_RACPrimaryRACPhysicalStandby.pdf]
  
  According to the documentation, all four instances of sleep should receive repeat primary, but that one instance should apply the redo (managed recovery). For some reason, only 3 of my four bodies receive remake and one (of these 3) performing recovery manged. The fourth instance is not do anything at all. I manually archived logs on all the main bodies several times to try to see any activity on this instance of Eve an idle.
  
  Issues related to the:
  1. How can I start receives data from Redo on this idle instance?

  In my view, it is possible to perform the configuration described in the MAA document strictly, (receiving multiple instances, 1-> 1, 2, -> 2, 3--> 3, 4-> 4), but that would require some artisanal dynamics of the addresses on the fly.

  You certainly don't want to hardcode your tns to include only one of the instances of Eve as a receiver.

  On the other hand, it is possible to include all 4 addresses instance ensuring the definition of DSTAGE tns in each of the primary tnsnames.ora but we're at the mercy of (client-side and server-side) load balancing, which does not guarantee the strict division of instance primary receiver 1 to 1.

  It's actually more robust to let the Oracle dataguard broker to manage its RAC-> dataguard RAC configuration.

  Interestingly, paper MAA Oracle does not subscribe to what are the infrastructure of the Oracle, process of dataguard broker, not done behind the scenes, when she puts in place of destinations-> RAC RAC configuration.

  In reality, that all primary instances get configured with the same service = "(description...

  "... and only the apply instance receives all of redo from all primary instances.

  When dataguard broker detects failure of apply instance, it dynamically resets all of the corresponding destinations on primary to a surviving standby instance and restarts recovery on it dynamically.

• DataGuard with other software

  Hi all

  9i

  RHEL5

  I configured dataguard physical standby in our primary database. We have therefore 3 database/servers of PROD, STBY, REPL, that is being configured.

  Then 3rd is configured to "replication" from primary to a different database replication replace. is between PROD and REPL only.

  The tool/software replication can the the following:

  1. materialized views

  2 advanced replication

  3. flow

  4. Golden Gate

  5 dbvisit

  My question is, if I switch to STBY, what happens to my replication system?

  Thank you very much

  JC

  Below the document explains the methods of replication in DG Environment

  Golden gate:

  Support for GoldenGate extraction to Active Data Guard again? (Doc ID 1299805.1)

  Streams:

  Data Guard & Streams

  Materialized views:

  Materialized views and Physical standby database (Doc ID 1265625.1)

  How materialized views are managed in a Data Guard environment (Doc ID 816863.1)

• OEM and dataguard

  Hi all

  11.2.0.1

  We have existing database of PROD with installed OEM (generic local).

  If I configure dataguard physical standby for her, he will also replicate the OEM to the physical expectations? or I need to reinstall on it?

  Thank you all,

  pK

  Hi pK,.

  It is not possible to monitor and administer a basic physical or logical standby, IE using Enterprise Manager Database Control Data Guard.  This is mainly due to the fact that Database Control is designed to monitor the 1-database and an environment Data Guard, by definition, includes more than 1 database.

  If you attempt to run emca against a database of pending, you will get an error such as ORA-01219: database is not open.

  Of course, database Control, can be used to monitor the current main database (with no capacity to administer or control Data Guard related features).  In such a case, failover Database Control needs to be reconfigured to run on the new primary database using the commands described in detail in Note 278100.1 how to remove, create and recreate DB Control In A Database, section c. recreate/ReConfig DB control, Option 2 10 g. recreate the control DB Configuration files and repository.

  Enterprise Manager Grid Control or EM 12 c Cloud Control provides the functionality for visualization, monitoring and administration of primary databases and standby in a Data Guard configuration.

  REF

  It is Possible to configure the database for a logical or physical Standby Database command? (Doc ID 315116.1)

  Kind regards

  Rahul

• DG 1 physical standby

  Hello

  11.2.0.1

  We have a DEV database which is also as critical as the PROD db since there are many strategic projects running on it.

  This DEV database has been configured with dataguard physical standby on the DR server.

  Well, that's what I read in the documents only as I'm newly taken over as a dba. Setup has been done outsourced

  by 3rd party, and I can't talk to the implentors, as they were long gone.

  When I check the DR server that only the PROD is running and the DEV was arrested for the last 12 months.

  My question is if the physical standby is stopped for almost a year, can it still be activated? or is it still valid?

  This will affect the performance of the primary database? Since he could always check heartbeat for her? or delay some process as archivers?

  How to disable the physical standby from the primary database?

  Thank you

  Petra k.

  Hello;

  "If the physical standby is stopped for almost a year, can it still be activated? or is it still valid? »

  No and no. Even a tiny production database would have generated and remove too many archives in a year for the old standby at any value. Rebuild it.

  How to disable the physical standby from the primary database?

  You could do just that: (primary side)

  ALTER system set log_archive_dest_state_2 = reporter;

  And you can clean all the parameters for DG too.

  How to remove a primary database data protection Configuration (Doc ID 733794.1)

  Best regards

  mseberg

• DataGuard Vs Standby

  Hi all
  
  I know that most of you have lived in the implementation of the eve and dataguard. For me I just did on the theory of testing and not yet on production because
  It generats a lot of newspapers and also affects performance.
  For me I just also stop and copy the files to another drive, since we can afford to stop the database after office hours.
  I didn't think that the database may crash at any time of the day and I'll lose transactions of the day.
  But how to protect your data which is in archivelog mode when the itelf disc where newspapers get crashed and becomes unrecoverable.
  And the same copy reduntant to roll forward are also on the same raid drive?
  
  What were you doing when this happens the of the archivelogs is missing or has a size of 0 bytes (meaning got corrupted)?
  
  Thank you very much
  
  Yxz

  Hello;

  If you use a Standard edition, you can configure Data Guard a poor man. There is a note about it here:

  http://managingoracle.blogspot.com/2012/04/Oracle-manual-standby-no-DataGuard.html

  You can switch the TNSNAMES if primary plant as an option.

  If you are not really you should fix this ASAP in archivelog mode.

  Best regards

  mseberg

• Script to remove applied connects on the physics of the standby database

  Hello
  
  Our prod DB is configured with Dataguard physical standby Database (10.2.0.4) on Linux.
  
  I'm trying to write a script to delete the archives that are applied on a standby database to DR (standby server).
  
  Any help will be appreciated.
  
  Kind regards
  Viswa08

  viswa08 wrote:
  Hi pavan,.

  I received the script of OTN itself below is the script and I can't understand how the sequence is deleted.

  ******************************************************************************************
  ###################
  #! / usr/bin/ksh

  script = ' basename $0'.

  SID = RW2P
  dir = / u04/oracle / $sid/arch

  Prof = $HOME /'echo ${sid} | AWK ' {print tolower($0)} ".profile
  tmpf=$home/.$script.tmp

  LogCount = 100

  function GetAppliedLogfileSequenceNumber
  {
  sqlplus-s/nolog $tmpf
  Connect / as sysdba
  Go head
  set pages 0
  Select max(sequence#) from v$ archived_log in case of application = 'YES ';
  output
  EXPRESSIONS OF FOLKLORE
  return
  }

  If [f $prof]
  then
  . $prof
  on the other
  print ' $script: $prof no such file or directory ".
  Output 1
  FI
  If [d $dir]

  If there is the $dir

  then
  CD $dir

  change directory to $dir

  GetAppliedLogfileSequenceNumber

  call the GetAppliedLogfileSequenceNumber function

  So [s $tmpf]

  If there is no sequence max

  then
  Count ='cat $tmpf | AWK '{print $1} ' | SED 1 d; »

  Set as maxsequence

  If [${#count} - No 0]

  If the counter is not zero

  then
  Let count = $count - $logcount

  Assign County County-100 the script does not remove the last 100 archivelogs applied.

  If ((County<=>

  If count is less than or equal to zero

  then
  print ' $script: journal count is set to zero (or not) no or newspapers to remove.
  Exit 0
  FI
  on the other
  Exit 0
  FI
  on the other
  print ' $script: no archive or newspapers to remove.
  Exit 0
  FI
  While [http:// f ${sid} _arch_1_$ {number} .arc]

  If $ORACLE_SID_arch_1_count.arc is

  do
  RM-f ${sid} _arch_1_$ {number} .arc

  delete

  RCODE = $?

  RCODE is assigned as the return code of the rm command.

  If ((rcode! = 0))

  If the return code is success

  then
  Print "$script: can't delete: ${sid} _arch_1_$ {number} .arc.
  Output 1
  on the other
  Let count = $count - 1

  reduce the number of 1

  FI
  fact
  RM-f $tmpf

  delete the temporary file.

  on the other
  print ' $script: $dir no such file or directory ".
  Output 1
  FI
  Exit 0

  Hope will help you.

  Kind regards.

• Redologs mandatory standby

  Hello
  
  I am new to the data protection feature.
  I have create a standby physical and everything works perfectly. The archives are shipped and applied.
  Some time, I set standby mode only the morning and evening reading back into recovery mode.
  
  I just need one point to delete.
  I have not any redolog Eve and everything works fine. It sounds not so important to have?
  Is this such a livable system? I mean that I intend to use my watch for a read only database to put the part of the load off the primary (for purposes of statistics).
  
  
  Work on AIX with Oracle 10.2
  
  Thank you

  Hell'o,.

  Newspapers waiting to roll forward is not mandatory.
  They are useful if you plan to do active data guard and a failover, actions of the digital switchover and other characteristic numbers of guard.
  It is also useful if you have cascading to the standby database.

  You must then create the redolog watches over the database pending and you can also create some on the primary database to be able to make the switch from one side to the other and return to... More like a preventive action.

  Note: When you create redolog Eve, don't forget to create a group more then the number of groups of primary redo logs (good practices).

  If you use your just sleep at the end you speak of, then you can leave it all as if it was without any problem.
  Just think you will need the redolog Eve the day you use your Eve fine dataguard more advanced.

  Loïc

• The change is not propagated to sleep

  DB version: 12.1.0.2 (I think that this applies to the 11.2.0.4 as well)

  OS: OL 7.2

  Configuration: Active dataguard + physical standby in maximum availability mode

  ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

  In my DB server test awaiting, I cancelled recovery Managed for something.

  -I haven standby DB

  SQL > alter database recover managed standby database cancel;

  Database altered.

  - Then I inserted a registration in primary education in a test table.

  -In primary DB

  SQL > select max(sequence#) from v$ archived_log;

  MAX(SEQUENCE#)

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

  20

  SQL > insert into scott.mumbo values (836);

  1 line of creation.

  SQL > commit;

  Validation complete.

  SQL > select * from scott.mumbo;

  EMPID

  ----------

  3

  76

  87

  39

  29

  836

  6 selected lines.

  SQL > select max(sequence#) from v$ archived_log;

  MAX(SEQUENCE#)

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

  20

  Pending side, I opened the DB (for active data guard) and started recovery managed real-time apply mode.

  But the new record (836) that was inserted in the primary scott.mumbo table is not propagated in standby mode. Primary and secondary sequence number.

  -In standby DB

  SQL > alter database open;

  Database altered.

  SQL > ALTER DATABASE RECOVER MANAGED STANDBY DATABASE with the HELP of DISCONNECTING from file LOG CURRENT;

  Database altered.

  SQL > select * from scott.mumbo;

  EMPID

  ----------

  3

  76

  87

  39

  29

  SQL > select thread #, PROCESS, STATUS, THREAD #, SEQUENCE #, BLOCK #, BLOCKS OF V$ MANAGED_STANDBY;

  THREAD # PROCESS STATUS THREAD # SEQUENCE # BLOCK # BLOCKS

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

  0 ARCH. CONNECTED 0 0 0 0

  0 ARCH. CONNECTED 0 0 0 0

  0 ARCH. CONNECTED 0 0 0 0

  0 ARCH. CONNECTED 0 0 0 0

  0 RFS       IDLE                  0          0          0          0

  1 APPLYING_LOG 1 21 759 102400 MRP0

  6 selected lines.

  SQL > select max(sequence#) from v$ archived_log in case of APPLICATION = 'YES ';

  MAX(SEQUENCE#)

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

  20

  The status of the MRP shows APPLYING_LOG and waited 20 minutes. No idea why MRP is not apply changes?

  -At the level of the backup server operating system, I can see running mrp

  [oracle@urcfnth187 ~] $ ps - ef | grep mrp

  Oracle 9729 1 0 00:43?        00:00:00 ora_mrp0_MTCFMS_DR0

  After waiting for 15 minutes, I changed the log in the primary and change has been propagated to sleep and now I could see the 836 folder in the test table.

  But in real time to apply, roll forward is directly read redo buffer and sent to the site of relief and appllied it. The change must therefore spread forward without a log switch. Right?

  Apply real-time applies directly from the SRL. In your case, MRP did not work on standby when insert you the record and committed the primary.

  If you use the availability of Max, it again is not instantly written as MRP was not running, then it would behave as if in performance max mode and committed on the primary record. Once you have started the MRP and newspaper under the primary voltage, it again has been shipped and written in standby mode.

  Update:-I would like to be more specific before you start to get in the comments of misunderstanding. In your case, the new record was not visible on standby may be due to network problem with primary cannot send it again or sleep impossible to fetch it. That's why a switch to journal in your case, on the primary with force shipped and applied it again to wait.

  -Jonathan Rolland