Oracle hot backup confusion

Similar Questions

• Where is the official Oracle documentation that describes the hot backup?

  I used the cruelly utility failure to search for Metalink and Google, but the closest thing, I found is a description on "ask Tom".
  
  I'm looking for official Oracle documentation that describes what is happening internally during a hot backup. There are many websites and blogs out there that describe, but I'm looking for a description of Oracle themselves. The only things I found in the Oracle documentation until now is to know how to start and stop the hot backup, but not what is happening internally. GR 11, 2 preference if it made a difference.
  
  Thanks in advance.

  Hello;

  Its probably gone RMAN covers this.

  The closest would be:

  Backup and recovery reference 11g Release 2 (11.2) E10643-06

  Or

  Backup and recovery User Guide 11 g Release 2 (11.2) E10642-05

  See Chapter 8 - Concepts of backup RMAN - online backups and backup Mode

  Best regards

  mseberg

• hot backup managed user problem

  Jin
  
  A few days ago, I took the managed user my database hot backup. And then, to check if the data file are in correct mode, I ran the query:
  
  SQL > select * sauvegardΘ v$.
  
  FOLDER # STATUS CHANGES # TIMES
  ---------- ------------------ ---------- ---------
  1. NO ASSETS 6302304272 23 JANUARY 10
  2. NO ASSETS 6302305573 23 JANUARY 10
  3. NO ASSETS 6302304259 23 JANUARY 10
  4. NO ASSETS 6302305573 23 JANUARY 10
  5. NO ASSETS 6302305573 23 JANUARY 10
  6 0 NO ASSETS
  7 0 NO ASSETS
  
  
  
  And, I didn't know that all data files have completed end backup mode. But today, when I query v$ datafile_header to check the column fuzzy then I was shocked to see the result below:
  
  SQL > select fuzzy from v$ datafile_header;
  
  FUZ
  ---
  YES
  YES
  YES
  YES
  YES
  YES
  YES
  
  7 selected lines.
  
  I use oracle 9.2.0.8 on windows server 2003.
  
  
  Although the view of backup $ v shows the same 'NOT ACTIVE' because all data files.this make me very confused.
  
  
  Please give me some feedback...
  
  Kind regards

  V$ BACKUP and V$ DATAFILE_HEADER are two different views. Why confuse yourself? They are there to meet different requirements.

  V$ BACKUP is used to check if a data file is in backup mode.

  V$ DATAFILE_HEADER is used to check the header of a data file. Of course a datafile is BLURRED when the database is OPENED - independently of the question whether an online backup is running or has run earlier.

  Hemant K Collette
  http://hemantoracledba.blogspot.com

• Closed APB hot backup

  Hello

  Could someone please help me to clarify the following:

  

  The above is the new guide features 12 c University student.

  Is not "alter database begin backup" called "backup of user management?

  "Hot backup" AFAIK means online of RMAN backup and requires that the database is in log mode archive.

  Also the instructer said "If the PDB is closed, you take a hot backup, that APB is not save, because the files are not in line.» "If the file is not open to the database, RMAN cannot hot backup.

  I just test rman backup in 12.1.0.1 when the PIC is opened and it included also all the PDB, including those that have been closed and in mounted condition. So what is the correct history?

  Thank you!

  Unfortunately, I can ask is no longer the instructor, but I noticed yesterday that I have actually ask your instructor about the same question a year ago. The answer was however really does not correspond to the question. In any case I guess it was just some confusion about the context.

  From what I understand, there are two ways to make a hot backup. One is called "user managed backup" by putting the database in backup mode by using the "alter database begin backup" statement and the other is to use RMAN. Two hot backupl methods require that the database is in log mode archive, but RMAN, by default, unlike "user managed backup" includes storage in offline mode.

  This means while a database is hot backup mode using "user managed backup", cannot close a PDB file and doing so will trigger an error ORA-01149. RMAN however has no such restrictions and will automatically include the backup of the PDB closed.

  I hope that my conclusion is correct and I am considering this thread supposed replied.

• Cold Oracle EBS backup and restore does not work

  Hello
  
  I made a backup with tar, while the database is open and not in mode archive (archivelog disabled) and restored on another machine.
  OS = AIX 5.3
  Oracle EBS 11.5.9
  9i database
  
  When I started the database I had errors.
  
  Is there a metalink document that will help me to recover the database?
  
  
  I followed the following Metalink:
  
  Database startup fails with ORA-01113, ORA-01110 errors [146039.1 ID]
  Common causes and Solutions in case of error ORA-1113 in backup & recovery [ID 183367.1]
  However, I note that the foregoing applies in the case of archivelogs = active
  My case: archivelogs = disabled
  
  SQL > startup mount pfile=/oracle/proddb/9.2.0/dbs/initPROD.ora
  ORACLE instance started.
  
  Total System Global Area 615482512 bytes
  Bytes of size 742544 fixed
  436207616 variable size bytes
  167772160 of database buffers bytes
  Redo buffers 10760192 bytes
  Mounted database.
  SQL > SELECT V1. FILE NO., NAME
  V $ BACKUP V1, V$ DATAFILE V2
  WHERE V1. STATUS = 'ACTIVE' AND V1. FILE # = V2. FILE NO.; 2 3
  
  no selected line
  SQL > restore database using backup controlfile until cancel;
  ORA-00279: change 3368247542 September at 10/02/2013 03:07:21 needed for thread
  1
  ORA-00289: suggestion: /oracle/proddb/9.2.0/dbs/arch1_637.dbf
  ORA-00280: change 3368247542 thread 1 is in sequence #637
  Specify the log: {< RET > = suggested |} Filename | AUTO | CANCEL}
  Cancel
  ORA-01547: WARNING: RECOVER succeeded but OPEN RESETLOGS would get below error
  ORA-01194: file 1 needs a recovery more match
  ORA-01110: data file 1: ' / oracle/proddata/system01.dbf'
  ORA-01112: media recovery not started
  
  SQL > restore database using backup controlfile until cancel;
  ORA-00279: change 3368247542 September at 10/02/2013 03:07:21 needed for thread
  1
  ORA-00289: suggestion: /oracle/proddb/9.2.0/dbs/arch1_637.dbf
  ORA-00280: change 3368247542 thread 1 is in sequence #637
  Specify the log: {< RET > = suggested |} Filename | AUTO | CANCEL}
  
  ORA-00308: cannot open archived log ' / oracle/proddb/9.2.0/dbs/arch1_637.dbf'
  ORA-27037: unable to get file status
  IBM AIX RISC System/6000 error: 2: no such file or directory
  Additional information: 3
  ORA-01547: WARNING: RECOVER succeeded but OPEN RESETLOGS would get below error
  ORA-01194: file 1 needs a recovery more match
  ORA-01110: data file 1: ' / oracle/proddata/system01.dbf'
  SQL > alter database open resetlogs;
  ALTER database open resetlogs
  *
  ERROR on line 1:
  ORA-01194: file 1 needs a recovery more match
  ORA-01110: data file 1: ' / oracle/proddata/system01.dbf'
  
  SQL > select HXFIL File_num, substr(HXFNM,1,40) filename, Type FHTYP, HXERR validity.
  FHSCN SNA, FHTNM TABLESPACE_NAME, status FHSTA, FHRBA_SEQ sequence
  x $ KCVFH;
  
  .....
  .....
  .....
  
  419 selected lines.

  981553 wrote:
  I understand that hot backups can be done with noarchivelog and data will be inconsistent.

  Fix.

  However, the database should not be able to recover some data are there in data files and commissioning normally.

  This is my question and my concern. I looked for Metalink and couldn't find anything about it.

  I want to start the database with all this that data is present after the backup. All the solutions for such a scenario?

  If you took an offline backup, you can recover the database to restore all the files and bring services up (without recovery is needed here).

  And, as you mentioned "hot backups can be done with noarchivelog and data will be inconsistent.

  Thank you
  Hussein

• Hot backup question

  Our DBA functions are currently outsourced, and I'm just curious about something.
  
  We asked them to copy a schema from our database of Production (during the day) to its equivalents in the acceptance and Test environments. We found a problem in that the maximum value of a primary key on a table was 100121, and the sequence used to generate the value of the primary key is only 100081.
  
  Is a risk to a backup in hot and any annex a backup must be taken after hours? Or maybe did something wrong?
  
  
  Oracle Database 10g Enterprise Edition Release 10.2.0.4.0 - 64bi

  If the request was to copy only a single schema, then it is unlikely that a "hot" backup would have been used. Most likely exp/imp or expdp/impdp would have been used to export the Production schema and import it into the other two databases.

  The sequences are always 'a problem' when export - import is used when the database is active. Export of the sequence and the export of the table (or tables) using the sequence are not at the same time / SNA so that the values can differ as they have updated in the source database.
  The plan is to restore order in the target database by manually incrementing until it reaches the desired target (100122).

  Hemant K Collette

• Hot backup

  We take the backup hot at 18:00 (daily, filled in 1-2 hours).
  At 23:30 all archive log files will be separately BackupExec backup and the log files are deleted from the file archive of the Oracle.
  
  Day before yesterday, for some reason, the copy of the data file hung up (after alter database begin backup has been published in the Hot backup script)
  That the copy has been hung up, 'alter database end backup' has been issued, the PB was in backup mode.
  The average while the deleted BackupExec database journal archive file Archive of the Oracle.
  
  Yesterday hot backup was completed successfully, only alter database begin backup failed (ORA-01146: cannot start online backup - folder 1 is already in the backup)
  After the hot backup 'alter database end backup' was published successfully.
  
  Given that logs archiving failed whien database is always in backup mode, will all the data generated duing the applied to the data file backup mode?
  
  Please notify.
  
  Thanks in advance,
  
  Arul

  You hear all changes (including the dml, ddl) applied to the database during the backup mode...

  YES.
  See http://www.bluegecko.net/oracle/oracle-tablespace-hot-backup-mode-revisited/

  Hemant K Collette

• When use REUSE/SET, no.-ARCHIVELOGS in create controlfile in HOT BACKUP?

  I am an Oracle DBA intern and I have the following queries. Good response with detailed explanations, I want to get my deleted concepts!
  
  T1 > > while a user managed hot backup, when we create a path for recovery (CREATE CONTROLFILE) control file when we use the control file created with the following options:
  * 1. REUSE / SET *.
  * 2. ARCHIVELOGS / NOARCHIVELOGS *.
  
  T2 > > in what scenarios re - create the control file while recovering the data from a hot backup files?
  
  Thanks a ton!
  Kind regards
  Blabla

  Specify REUSE to indicate that existing control files identified by CONTROL_FILES initialization parameter can be reused and overwritten.

  SET the DATABASE to change the name of the database.

• UNDOTABLESPACE and TABLESPACE TEMP in cold or hot backup database

  Hello
  
  I have a BACKUP, including logs generated after the backup archive. What is the use of the UNDOTABLESPACE and TABLESPACE TEMP datafile data file backup? When it is used?
  
  
  Kind regards
  
  007

  >
  I have the archives which have all entries are committed in it. Undoinformation is necessary for the restoration and recovery of the database? Can u explain more clearly?
  >

  The archives are as congested entries of redo logs. And the redo logs contain all changes to the database. Information cancellation will be there in the segments of cancellation that there can be some uncommitted transactions active in the segments of cancellation. Cancellation information is necessary to put the database in a consistent state. It is one of very important storage similar to the control file and system tablespace to bring a database

  Edited: Imagine a scenario. You make a backup hot at 07:00. And there are a lot of transactions happening before and during hot backup. For ex, I run a massive update to a table. This will cause the recovery to write logs of recovery that will be eventually written to logs archived. Now, you finish the backup mode. But I have not validated or cancelled my transaction. Say to the reasoning, I roll back the transaction.

  Now, you restore the backup. Do you think that my transaction will be in inconsistent state? Redo logs and archiveed newspapers had my transaction changes/redo entries. If you are not using undo in recovery segments, what will happen to my operation that I rolled back? Without cancel, half of my transaction (even if they were not truly committed) will be sitting in the database. This is where undo tablespace helps us in rolling back the transaction during your recovery. Remember, the cancellation information "peuvent" not all in recovery if you do not a complete database recovery. But Oracle should know that. If you need to undo tablespace in your backup set.

  I hope it's clear now

  Published by: VenkatB on Sep 15, 2011 10:17

• Hot backup maintained by users script? 11g (Unix) good measures?

  I'm looking for a script to hot backup maintained by users for GR 11, 2 under Unix. My client does not want to use RMAN.
  
  I think that the main steps are:
  
  -alter system switch logfile
  -change the system log archive current
  -alter database begin backup
  -alter database backup controlfile (and also the backup on the track)
  -copy the data files, archives, remake, controlfiles to another location
  -take the database out of hotbackup mode
  -J' have also capture spfile and other config files
  
  Recovery is just restore to the original location, at STARTUP and backup of the end.
  
  Basically, there may be cases where we have archives that have been made after the hotbackup we want to apply. In this case, it is RECOVER database with the HELP of BACKUP CONTROLFILE until CANCEL in order to apply the archives.
  
  His right?

  After you take the database out of Hot Backup, you must switch logfile (preferably several times) and backup generated archivelogs.
  You can back up archivelogs while the database is in backup mode but you must also back up the first archivelog (s) after the backup of the end.

  Hemant K Collette

  PS: Oracle provides documentation.
  http://download.Oracle.com/docs/CD/E11882_01/backup.112/e10642/part_umb.htm#BEHBHCBA

  Hemant K Collette

• Rebuild the basis of Standby hot backup existing

  Hi all
  
  I'm rebuilding the database of existing sleep using the current hot backup because of the huge gap between the primary and Standby, which cannot be restored by backup team
  
  Can someone please explian the steps to perform this activity.
  
  Oracle-9i version
  
  The size of the database - 50 GB
  
  Hot - Latest backup
  
  Kind regards
  RAM.

  Follow the same process as that used to create the eve of the first.

  http://download.Oracle.com/docs/CD/B10501_01/server.920/a96653/create_ps.htm#67525

  Skip the steps that should not be remade - for example to create the parameter file for the standby, the configuration of the listener and tnsnames, entries that you already have, etc.

  Hemant K Collette

• Need help with RMAN archive destination error during hot backup

  Hello
  
  Oracle 8.0.6, Win 2000 Server.
  
  We take the cold daily backup as our backup strategy. We wanted to implement the backup hot. If one day, I turned on the log mode archive and set the directory to archive in F:\drive to store all archived generated. I took the backup and put the database in noarchive log mode. We continued with offline backup.
  
  After a few days, I wanted to test the hot backup again. I created the directory of the archive in E:\archive\ instead of F:\archive\. When I started the hot backup, all files have been copied except the archives.
  
  The error was:
  
  RMAN-03022: compilation command: sql
  RMAN-06162: SQL statement: alter system archive log current
  RMAN-03023: executing command: sql
  
  RMAN-03022: compilation command: backup
  RMAN-03026: recovery error release of channel resources
  RMAN-08031: output channel: c1
  RMAN-08031: output channel: c2
  RMAN-00569: = error message stack follows =.
  RMAN-03002: failure during the compilation of command
  RMAN-03013: type command: backup
  RMAN-06089: journal archived F:\ARCHIVE\ARCH1_0000327116.ARC not found or at odds with catalog
  
  Complete recovery manager.
  
  RMAN is looking for the destination of the journal archive old i.e. 'f:\archive\ '. I don't have the backup of the old archives of destination. I put the new destination to archive E:\archive\ in the initorcl.ora file, rman is always looking for destination F:\archive\.
  
  Please suggest what to do. How can I remove old information from archive of controlfile and say rman, look in the new archives destination i.e. E:\archive\

  Hello
  WOW 8.0.6.
  Maybe you can run a cross?

  change archivelog than any overlap;

  Update
  Found some docs online for 8.0
  http://download.Oracle.com/docs/CD/A58617_01/server.804/a58396/TOC.htm
  She does not mention the overlap.
  With the change command, you should be able to do the cleaning as well.
  If you have moved the archivelogs to the new location, you can use catalog according to the doc.
  Kind regards
  Tycho

  Edited by: tychos on mrt-27-2011 08:26

• More redo generation during hot backup

  Hi all
  I want to know more over the generation of redo during hot backup.
  I know that in this case the data file headers are frozen. But I want to know how the whole thing goes. All did this experiment practically?
  Did any problems of performance because of this?
  
  
  
  Thank you
  Rafi.

  Hmm well, it basically means like this, assuming you have a block filled with 1s like that,

  ----
  1
  ----
  1
  ----
  1
  ----
  

  And this is the copy of the block that was sent to the backup by you and was taken over by the command cp o/s. While the cp command will make a backup of the block as it can be read much more small piece that, suppose that he would start with the 1st part of the block and starts copying to the o/s somewhere. At the same time feel, DBWR peaks upwards again and written on block 2, like that.

  ----
  1
  ----
  2
  ----
  1
  ----
  

  It can from the flie is always online and is available to be used by the database. Now, you have a problem. The image you sent for backup isn't th that is here now. Next time what cp command would return, he would find 2 instead of one in the 2nd part of blok and copy it since it is a mute command when comes to oracle. So now that you have in what you wanted in the backup and the backup are not the same images. You have a cracked now block which is of no use for the recovery. To this end, in the case of the hot backup, the first image of the block time is copied and is maintained. This saves us from the State to have 2 there. We have all the 1s who sent us. Now, if in the case, dbwr also changes the block, the original help oracle reover the block properly.

  HTH
  Aman...

• stop the database when the HOT backup running

  Hello
  
  Can someone pls tell... WHT can impact on the database if the database will get stop accidentally during the hot running backup...
  
  Thank you

  There will be no impact on db; because its ORACLE database; very smart and easy to use.

  What happens if the instance goes down for the tablespaces into backup mode?

  When you start the database after that, Oracle will say that it requires recovery. It's because the RCS has been frozen and this is the expected behavior, because if you restore the file copied, it needs to be recovered. (and the only way that oracle has to set its value in the copy is put in the current file while it is copied)

  In this case, you can issue:

  ALTER DATABASE END BACKUP; ALTER DATABASE OPEN; to open the database.

  But your backup is not usable, it must do it again.

  Source: http://knol.google.com/k/oracle-begin-backup-end-backup#

  HTH
  Girish Sharma

• HOT backup database 10g express edition

  Dear Sir
  I use oracle database 10g express edition with 3.2.1 apex.
  
  I'm planing to db.here hot backup entire database is 24 x 7 running... fashion archive
  Info
  
  pFile, control, password file
  System datafile
  SYSAUX datafile
  Undo tablespace file
  temporary tablespace file
  redo log files
  application data tablespace file
  application index tablespace file
  
  for backup of the entire database HOT including apex WORKSPACE... which tablespace or files are REQUIRED for backup?
  
  Thank you
  Engr.M.K Chowdhury

  Hello

  To be convinced, you can do the following:

  alter database begin backup;
  

  Then, you check the backup status of your datafilles with the following query:

  select file#, status
  from v$backup
  order by file#;
  

  If the status is ACTIVE, this means that the data files are in backup mode, and you can copy them.

  The redologs online, you can save them but this is not mandatory, because you have logs archived.

  It is important to get all archived logs generated during (and after) the backup.

  To do this, you can run the following right after the hot data files backup:

  alter system archive log current;
  

  It will generate a log archived that you need to include in your backup.

  If you lose all your data (including the redologs online), after you restore the data files, you need to retrieve
  the database. During this operation, you apply all archived logs and then you reset the redologs. It is a
  incomplete recovery.

  However, for the online redo logs, it is important to multiplex to another disk. So, if you lose members on
  a single disk, you can perform a full restore with their copy on the second disc.

  Hope this helps.
  Best regards
  Jean Valentine