Backup hot NOARCHIVELOG mode?

Similar Questions

• Backup/restore tablespaces in noarchivelog mode

  In our office, there is a box of oracle 10g that is used for testing purposes.
  He got only 1 database, but each user has its own tablespace.
  As it is not a 'live' box, archivelog and flashrecovery have been disabled by the dba likely in favor of speed.
  
  I have a suit of NUnit test that crosses a tablespace, and I'm trying to restore the tablespace before each race.
  In noarchivelog mode, RMAN allows me to take the tablespace offline and create a backup - well.
  Later, I'm capable of taking the tablespace offline again and issue a command to "RESTORE the TABLESPACE.
  But then I can't bring the tablespace online without issuance of an order to 'RECOVER '.
  I tried the NOREDO option but it still says that the file has a media recovery before it can be brought online.
  
  Is there a way to bring the tablespace online without applying REDO logs?
  Backup/restore of tablespace or not possible because of noarchivelog?
  
  Thank you.

  He asks for any newspaper archive, when you turn the fire 'recover' command?

  When you take a tablespace offline, the tablespace is not updated with the oracle's internal clock. So when you try to bring the tablespace online, oracle estimated that the tablespace is not in harmony with the other storage spaces. That is why it requires recovery. If you turn the fire back the command simply updates the tablespace and brings online storage space after applying the data available in the online redo log.

  But if the data has been flushed to redo, then it is difficult to synchronize the tablespace.

  Kind regards
  S.K.

• Restore, recover the backup hot inevitably will be OPEN RESETLOGS

  DB version: 11.2
  Platform: Solaris / AIX
  
  
  
  Cold backup is performed when the DB is in editing mode, and when no recovery is generated during the backup.
  So when you restore this backup, you can open the database with

  alter database open;

  But when we take the backup hot of a DB, we will never be able to redo backup generated in restoration online logs when the hot backup. Thus, during recovery, we can apply only archived redo logs from the source DB and we are obliged to skip applying entries online to the DB of the Source again.
  Thus, hot backup recovery will inevitably end up by in OPEN RESETLOGS.
  

  alter database open resetlogs

  These assumptions are good?

  If you restore and recover to the same server with the current controlfile and remake online still present newspapers, you would be able to make a full recovery. In such a case it is not necessary to OPEN RESETLOGS.
  All other cases of hot backup are incomplete recoveries and require an OPEN RESETLOGS.

  In the case of a cold backup taken by RMAN, you should always an OPEN RESETLOGS because RMAN does not include online of Redo Log files.
  http://hemantoracledba.blogspot.com/2008/05/RMAN-cold-consistent-backup-and-restore.html

  Hemant K Collette

• DataFile recover in the event of noarchivelog mode

  my database is in noarchivelog mode and there is no backup available, and one of my datafile is currupted so how can I recover the data file?
  
  
  Thanks in advance

  Julien Chevallier wrote:
  my database is in noarchivelog mode and there is no backup available, and one of my datafile is currupted so how can I recover the data file?

  Thanks in advance

  Version what ever. To retrieve the archive data file must be activated. NOARCHIVE norecovery...

• The user managed backup hot

  Hello
  
  When we started a tablespace begin backup mode State, I know that the RCS data files will be blocked and the recovery log generation will be more. I just want to know where the DML activities occur when the tablespace mode backup how to get updated records? why there are no changes in the SNA? why there is excess redo generation compared with the normal DML region?
  
  Vincent

  This is not true, as a simple audit would have told you.

  SQL> select file#, checkpoint_change# from v$datafile;
  
       FILE# CHECKPOINT_CHANGE#
  ---------- ------------------
           1         7667692204
           2         7667692204
           3         7667692204
           4         7667692204
           5         7667692204
           6         7667692204
           7         7667692204
           8         7667692204
           9         7667692204
          10         7667692204
          11         7667692204
          12         7667692204
  
  12 rows selected.
  

  Note: we start with all files with the same checkpoint change number. All data is synchronized

  SQL> alter tablespace data_medium begin backup;
  
  Tablespace altered.
  
  SQL> select file#, checkpoint_change# from v$datafile;
  
       FILE# CHECKPOINT_CHANGE#
  ---------- ------------------
           1         7667692204
           2         7667692204
           3         7667692204
           4         7667692204
           5         7667692204
           6         7667727909
           7         7667692204
           8         7667692204
           9         7667692204
          10         7667692204
          11         7667692204
          12         7667692204
  
  12 rows selected.
  

  Note: file 6 is now out-of-synch with the rest of the database, as it has been in hot backup mode. Change of control point number is ahead of the rest of the database.

  SQL> alter tablespace data_medium end backup;
  
  Tablespace altered.
  
  SQL> select file#, checkpoint_change# from v$datafile;
  
       FILE# CHECKPOINT_CHANGE#
  ---------- ------------------
           1         7667692204
           2         7667692204
           3         7667692204
           4         7667692204
           5         7667692204
           6         7667727909
           7         7667692204
           8         7667692204
           9         7667692204
          10         7667692204
          11         7667692204
          12         7667692204
  
  12 rows selected.
  

  And to note now that even if the END BACKUP command has been issued, the data files remain in exactly the same state that they were ahead. File 6 is always out of sync with all other files. No additional control points have been issued, and nothing further happened to produce 6 or anything else to do things to re - sync.

  Put a tablespace IN backup mode hot indeed causes a control point will take place for this table space. But out of hot backup mode does NOT work. The SNA in the headers of this tablespace will be re - synchronize the next time there is a checkpoint of the whole system, but not until then.

  So no, it's NOT ' updated when we take tablespace in backup mode using online redo log files "-Although the syntax of this sentence is so weird as to defy rational explanation and require pure conjecture.

  As for questions of the original poster, so:

  I just want to know where the DML activities occur when the tablespace mode backup how to get updated records? why there are no changes in the SNA? why there is excess redo generation compared with the normal DML region?

  1 DML activities occur quite normally. Put a tablespace in backup mode hot stops CKPT update its header, but does absolutely nothing to prevent your server to update another process the contents of the file. If you update every row to a million lines of table during a hot backup of the repository containing this table, each row is updated, exactly as he would have done if no hot backup took place. It is not property of a backup hot, after all, if the content data file was unflappable frozen for the duration, would it?

  2 there IS a change in the SCN when you put a tablespace in mode backup hot. See above. When you exit the hot backup mode, there is no special update for the RCS... but the next time that all the headers of data file are updated through a checkpoint at the system level (caused by things like a log switch or the ADMINISTRATOR requesting a) , then the header will be updated with the SNA as all the other data file headers.

  Get some more: this is the checkpoint change the number that is stored in the headers of data files, and it is that which does not change during a backup hot. Number of system change, including change of control point number is simply a snapshot at a point in time, increases continuously throughout the system, all the time.

  3. in a user managed backup, it's the operating system that copies things and it copies the things that the operating system is aware of--such as the "operating system block', which is commonly size of 512 bytes. An Oracle block consists of several operating system blocks (16 in the case of a block of 8 K and blocks of 512 bytes OS). Oracle knows 16 blocks BONE to form a block Oracle; the operating system itself has no idea of their interdependence.

  Therefore, the silent operating system will copy one of these 16 blocks at any time as seems it, because he realizes that if it copies one, she really should to copy other 15 institutions. If it were a cold maintained by users of backup, which is not serious, because nothing would be happening to all the 16 blocks of BONE, and as long as it took the BONES to copy all of them, the result would be the same, as if all 16 had been copied at the same time.

  But it's a hot backup, it happens at these blocks. Therefore, it is possible for the operating system to copy, say, three of them, just at the moment where Oracle is about to change one of its blocks with an update or insertion. Three blocks of BONE thus copied in one State and the other thirteen OS blocks copied in another State. You now have a block of 8K in a backup that is internally in all directions and totally unusable. It's called a cracked block, and that's what happens when a single entity (BONE) copy unconscious things the importance invested in this is it the copy by another (the Oracle database).

  The cure for a broken block is to have a clean block copy as it was before the updates that happened to her. This clean copy of the block, you can apply all redo them later generated who touched it. So it is what guarantees Oracle exist. If a tablespace is in hot backup mode, the first update that is subject to change an Oracle block causes this whole block to write in it again connects. There's your own copy. The copy that creates the backup operation is the copied data file can now be fractured or not as chance allows: he'll always be a good version of it stored in the recovery logs.

  However, an update that generated usually 200 bytes redo just caused all of a block of 8K to copy in the redo logs. And this increase in the generation of redo is a feature of backups maintained by users (and why you should use RMAN).

• put the offline data file when the database is in NOARCHIVELOG mode

  My question is when the database is in log mode Archive No. I am not able to take the database offline.

  When I tried in my computer I noticed set-aside.

  CASES1:

  SYS > alter database datafile 5 offline;

  ERROR at line 1;

  ORA-01145: offline immediately rejected at least media recovery enabled.

  case 2:

  SYS > alter database datafile 5 immediate offline;

  ERROR at line 1;

  ORA - 00933:SQL not correctly completed order

  CASE3:

  I tried the command drop alter database datafile offline 6; (in NOARCHIVELOG mode) and it is to show the same effect that alter database datafile offline 6; (in ARCHIVELOG mode).
  

  * In NOARCHIVELOG mode are we really drop the data file to use the offline data file? You please tell me the effect of the dropkeyword.

  Oracle protects you. Please review ARCHIVELOG and NOARCHIVELOG mode.

  When you take the database offline in NOARCHIVELOG mode:

  (1) the file is offline

  (2) the RCS in which you took the database offline will be is more to be in the online redo logs

  (3) at this point, you need to perform a media recovery to enable again the file online (ie. restore the database)

  So you can see why you shouldn't do that.

  The DROP Offline command does not actually remove the data file. It removes the control file file details and updates the offline data file. It can be used for purposes of recovery or when you actually plan on deleting data files.

• A physical database ensures running in noarchivelog mode

  Primary database: cluster 2 node rac, 11.2.0.2 RDBMS and GI
  Standby Database: single instance 11.2.0.2
  OPERATING SYSTEM: AIX 7.1
  
  Hello
  
  I'm trying to implement a Data Guard standby system. To make this simpler system to the admin, I want to run the day before in NOARCHIVELOG mode database. This expectation is used only as a mirror of the production (to be copied to a 3rd site of activity of separate user) and will never pass mode. Now another set of archive logs seems to be redundant to me and another point of failure, so I want to disable archiving. I wanted to ask if all the experts of DG on the forums have already had experience with this? Does anyone know if this is recommended? I searched the Oracle Data guard documentation online... it is said that the primary must obviously be in mode archive, but I don't see anything specific on the standby. Please let me know if anyone has experience with this.
  
  Thank you

  Hello again;

  If this is the case, then I have nothing to worry about here. Does that sound right to you?
  

  Yes.

  To do this:

  b) monthly, shutdown this system and copy it to a 3rd site to host a separate set of business activity
  

  I would consider this case RMAN Duplicate, you can avoid the closure.

  And for that:

   resync the "mirror" with production
  

  All you have to do if SEE the primaries forward, then ENABLE and Data Guard will catch up.

  -Will post a detailed example here in a moment.

  From the beginning of the year, they wanted to work on our main server room (power). I've done the switchover to the waiting, but I kept log_archive_dest_state_2 on the new primary series to DEFER. Then I stop
  all databases of the main site. At the end of the work two days later, I started the databases on the primary site in standby mode. Log_archive_dest_state_2 switched to ACTIVATE. While the databases were back in SYNC
  I did another pass and he did. I used the TAF, and none of the users even notice.

  Best regards

  mseberg

• Cloning time-consuming RMAN backup & hot

  Dear administrators,
  
  We do both RMAN & Hot backup cloning enemy that takes more time to restore the TEST instance data files. SCP taking too long to copy the server target due to its volume of data (1.5 to) and mounting of the drive of RMAN backups on test server and restore this drive taking more than 8 hours.
  
  I heard that some companies are using a device to change the copy backup hot high volume of source dest faster for clone instance easily with in a few minutes. What is the name of device to take copies of source hot backup and restore to the server target.
  
  Test server specifications
  Processor cores 8 with 32 GB of memory.
  
  Thank you
  S/N

  We have about 2.5 TB of data. It takes about 10 hours to synchronize the data.
  But everything depends on the EMC technology.
  in my previous company, synchronization takes about 1 hour to 900G of data.

• Not complete checkpoint in NOARCHIVELOG mode

  Hello
  This is the first time I see this. In my 11.1.0.7 development database on SOLARIS, I see checkpoint do not message in the log file alerts and my database is running in NOARCHIVELOG mode. Can any expert light on what why this warning is here even in NOARCHIVELOG mode?
  
  Salman

  871174 wrote:
  Hello
  This is the first time I see this. In my 11.1.0.7 development database on SOLARIS, I see checkpoint do not message in the log file alerts and my database is running in NOARCHIVELOG mode. Can any expert light on what why this warning is here even in NOARCHIVELOG mode?

  Salman,

  The error has nothing to do with the archive log or no mode of journal archive the database, but the size of the recovery of your log files and other factors. The error means that you are not able to control your last point again them pretty fast until it can be reused. What is the size of the recovery of your log files?

  Aman...

• problem in restoring the backup hot

  Hello
  
  I tried to restore a backup hot production server to the test server, but could not restore the same as he showed the mention below
  
  ORA-01589: must use the RESET the LOGS or LOGS NORESET option to open database.
  
  When we tried with the reset option to open we get error mentioned.
  
  SQL > alter database open resetlogs;
  ALTER database open resetlogs
  *
  ERROR on line 1:
  ORA-01113: file 1 needs media recovery
  ORA-01110: data file 1: / data01/proddata/SYSTEM01. DBF'
  
  
  When we tried to recover by using the recover command, we got the following error message
  
  
  ORA-00308: cannot open archived log
  \backup\archive\ARC00120.001'
  ORA-27041: could not open the file
  04002 - OSD: could not open the file
  S/O-error: (OS 2) the system cannot find the file specified.
  
  But the journal of current archiving in the production database is ARC00119.001' that we copied on the test server, but she asked the next number that has not yet been generated.
  
  
  Please find the hotbackup script that we use.
  
  
  HOST mkdir/backup/SauvegardeQuotidienne/hotbkpPROD170709 /.
  coil /backup/dailybackup/hotbkpPROD170709/hotbkp.log
  Date of the HOST
  Select to_char (sysdate, ' dd/mm/rrrr hh24:mi:ss') Start_Time
  Double;
  alter tablespace SYSTEM begin backup;
  HOME cp /data01/proddata/system01.dbf/backup/SauvegardeQuotidienne/hotbkpPROD170709. /
  HOME cp /data01/proddata/system02.dbf/backup/SauvegardeQuotidienne/hotbkpPROD170709. /
  HOME cp /data01/proddata/system03.dbf/backup/SauvegardeQuotidienne/hotbkpPROD170709. /
  HOME cp /data01/proddata/system04.dbf/backup/SauvegardeQuotidienne/hotbkpPROD170709. /
  HOME cp /data01/proddata/system05.dbf/backup/SauvegardeQuotidienne/hotbkpPROD170709. /
  HOME cp /data01/proddata/system10.dbf/backup/SauvegardeQuotidienne/hotbkpPROD170709. /
  HOME cp /data01/proddata/system06.dbf/backup/SauvegardeQuotidienne/hotbkpPROD170709. /
  HOME cp /data01/proddata/system07.dbf/backup/SauvegardeQuotidienne/hotbkpPROD170709. /
  HOME cp /data01/proddata/system09.dbf/backup/SauvegardeQuotidienne/hotbkpPROD170709. /
  HOME cp /data01/proddata/system08.dbf/backup/SauvegardeQuotidienne/hotbkpPROD170709. /
  HOME cp /data01/proddata/system11.dbf/backup/SauvegardeQuotidienne/hotbkpPROD170709. /
  alter tablespace backup end of SYSTEM;
  alter tablespace CTXD begin backup;
  HOME cp /data03/proddata/ctxd01.dbf/backup/SauvegardeQuotidienne/hotbkpPROD170709. /
  change the tablespace CTXD end backup;
  alter tablespace APPS_TS_TX_IDX begin backup;
  HOME cp /data03/proddata/a_txn_ind01.dbf/backup/SauvegardeQuotidienne/hotbkpPROD170709. /
  HOME cp /data03/proddata/a_txn_ind02.dbf/backup/SauvegardeQuotidienne/hotbkpPROD170709. /
  HOME cp /data03/proddata/a_txn_ind03.dbf/backup/SauvegardeQuotidienne/hotbkpPROD170709. /
  HOME cp /data03/proddata/a_txn_ind04.dbf/backup/SauvegardeQuotidienne/hotbkpPROD170709. /
  HOME cp /data03/proddata/a_txn_ind05.dbf/backup/SauvegardeQuotidienne/hotbkpPROD170709. /
  change the tablespace APPS_TS_TX_IDX end backup;
  alter tablespace APPS_TS_TX_DATA begin backup;
  HOME cp /data03/proddata/a_txn_data01.dbf/backup/SauvegardeQuotidienne/hotbkpPROD170709. /
  HOME cp /data03/proddata/a_txn_data02.dbf/backup/SauvegardeQuotidienne/hotbkpPROD170709. /
  HOME cp /data03/proddata/a_txn_data03.dbf/backup/SauvegardeQuotidienne/hotbkpPROD170709. /
  change the tablespace APPS_TS_TX_DATA end backup;
  alter tablespace APPS_TS_ARCHIVE begin backup;
  HOME cp /arch/proddata/a_archive01.dbf/backup/SauvegardeQuotidienne/hotbkpPROD170709. /
  change the tablespace APPS_TS_ARCHIVE end backup;
  alter tablespace OLAP begin backup;
  HOME cp /arch/proddata/olap.dbf/backup/SauvegardeQuotidienne/hotbkpPROD170709. /
  change the OLAP end tablespace backup;
  alter tablespace APPS_TS_QUEUES begin backup;
  HOME cp /arch/proddata/a_queue02.dbf/backup/SauvegardeQuotidienne/hotbkpPROD170709. /
  HOME cp /arch/proddata/a_queue01.dbf/backup/SauvegardeQuotidienne/hotbkpPROD170709. /
  alter tablespace APPS_TS_QUEUES end backup;
  alter tablespace PORTAL begin backup;
  HOME cp /arch/proddata/portal01.dbf/backup/SauvegardeQuotidienne/hotbkpPROD170709. /
  change the PORTAL end tablespace backup;
  alter tablespace APPS_TS_SEED begin backup;
  HOME cp /arch/proddata/a_ref01.dbf/backup/SauvegardeQuotidienne/hotbkpPROD170709. /
  HOME cp /arch/proddata/a_ref02.dbf/backup/SauvegardeQuotidienne/hotbkpPROD170709. /
  change the tablespace APPS_TS_SEED end backup;
  alter tablespace APPS_TS_SUMMARY begin backup;
  HOME cp /arch/proddata/a_summ01.dbf/backup/SauvegardeQuotidienne/hotbkpPROD170709. /
  change the tablespace APPS_TS_SUMMARY end backup;
  alter tablespace APPS_TS_INTERFACE begin backup;
  HOME cp /arch/proddata/a_int01.dbf/backup/SauvegardeQuotidienne/hotbkpPROD170709. /
  change the tablespace APPS_TS_INTERFACE end backup;
  alter ODM tablespace begin backup;
  HOME cp /arch/proddata/odm.dbf/backup/SauvegardeQuotidienne/hotbkpPROD170709. /
  change the tablespace MDG end backup;
  alter tablespace OWAPUB begin backup;
  HOME cp /arch/proddata/owad01.dbf/backup/SauvegardeQuotidienne/hotbkpPROD170709. /
  change the tablespace OWAPUB end backup;
  alter tablespace APPS_TS_MEDIA begin backup;
  HOME cp /arch/proddata/a_media01.dbf/backup/SauvegardeQuotidienne/hotbkpPROD170709. /
  change the tablespace APPS_TS_MEDIA end backup;
  alter tablespace APPS_TS_NOLOGGING begin backup;
  HOME cp /arch/proddata/a_nolog01.dbf/backup/SauvegardeQuotidienne/hotbkpPROD170709. /
  change the tablespace APPS_TS_NOLOGGING end backup;
  alter tablespace APPS_UNDOTS1 begin backup;
  HOME cp /arch/proddata/undo01.dbf/backup/SauvegardeQuotidienne/hotbkpPROD170709. /
  alter tablespace APPS_UNDOTS1 backup end;
  HOME cp /data01/proddata/cntrl01.dbf/backup/SauvegardeQuotidienne/hotbkpPROD170709. /
  ALTER database backup controlfile to trace.
  HOME cp /oracle/proddb/9.2.0/admin/PROD_gurlfiprddb01/udump/*.trc/backup/SauvegardeQuotidienne/hotbkpPROD170709. /
  ALTER database backup controlfile to ' / backup/dailybackup/hotbkpPROD170709/control.dbf';
  alter system archive log current;
  Select to_char (sysdate, ' dd/mm/rrrr hh24:mi:ss')
  End_Time
  Double;

  I said above that you can use the allowresetlogs_corruption initialization parameter and you can open the database.
  Try this way.

• Very hot battery Mod - while loading with phone wireless

  Has anyone else noticed that gets battery mod VERY hot while the load connected to the phone? Mine was powerful warm when I got him on the phone and plugged to the charger.

  Is this normal planned?

  There are circuits that are designed to disable loading if the temp goes out a certain range. They need people to avoid any risk of damage.

• How can I make a backup in safe mode?

  After HP installed a new motherboard in my computer it does not start in normal mode due to a conflict pilot somewhere they can't find.  But I can run it in safe mode.

  HP has decided to send me a new (and better) computer rather than trying to resolve the conflict.  But my problem is - no backup program does work in safe mode!  The thing Windows Easy Transfer does not work.
  I desperately need to get all of my Office 2010 settings and configurations moved to the new machine.  And ideally, I'd like to get all my programs and settings moved as well.  I have an external hard drive that has plenty of space for everything.
  I would like to make a backup image of August that contains all programs and settings, I need.  But the old machine is running Windows 7 Professional and the new a Windows 7 Home Premium and I do not know if the backup/restore functions are the same.  Anyone know if I can use the backup/restore of Windows in Windows 7 Home Premium to restore settings and programs to a Windows 7 backup professional?  If so, it would be what I need.
  If not, anyone know if there is a way to save these settings in safe mode and programs?
  (I can't believe none of this works in safe mode.)  When you need to run in safe mode, you are obviously in a critical State and it seems that would be a time that many people would need to make a backup).
  Thank you!

  Ah no!  I did not try it.  This would have been an excellent suggestion.

  That's what I hate about these forums... you get a useful answer, a week after you ask!  (not your fault, I know).

  But I'm already in place and running on the machine again and the former is on its way via FedEx to HP.

• restore user managed with RMAN backup hot

  Is it possible to restore hot backup with RMAN user managed and recover the database?
  
  What is use of command RMAN CATALOG?
  
  Oracle doc said: [http://docs.oracle.com/cd/B14117_01/server.101/b10770/rcmsynta11.htm#79814]
  
  (1) order catalog added a copy of the data file information, archived redo log or copy of control in the repository files.
  (2) add metadata to a copy of the data file managed by users, copy of control files, archived log or backup piece to the RMAN Repository.
  
  After using the 'CATALOGUE' to add metadata to managed backup in the catalog of recovery and after having made "RESYNC CATALOG", I still don't see my backup listed in the output of
  
  RMAN > list backup
  
  Published by: Gaurav Mhatre on April 23, 2013 10:35

  Gaurav Mhatre says:
  After using the 'CATALOGUE' to add metadata to managed backup in the catalog of recovery and after having made "RESYNC CATALOG", I still don't see my backup listed in the output of

  RMAN > list backup

  backup of the list won't show you the results.your not done backup with rman directly taken by data copy.so file

  RMAN > list copy;

  above cmd will show that you have desired results check it out.

• Veeam Backup and the Mode of free features

  People,

  I started a support with Veeam call after the installation of their new backup and replication 4.1.1 after getting an error "free ESXi is not supported" after some troubleshooting tech me determine what level of license Info Service is listed, ESXFull, or ESXBasic. After determining that it is ESBasic they advise that we are 'free' feature mode and call the VMWare licenses.

  I'm missing somethgin here, I thougth all ESXi were free, a matter that deals with this and how do I get this to ESXFull?

  Thank you

  Free ESXi is free. It can be licensed for use with the rest of the vSphere product suite. In free mode, there are several things that are not available, including the storage API. Veeam backup uses the storage API, which means that you must have a licensed version of ESXi. The simplest version of a bundle under license is http://www.vmware.com/vmwarestore/buyvsphere-small-business.html

• Question about backup hot with Tablespace with no logging

  Hi, we have an application that using a temporary tables (tables of work). The application maintains a persistent connection to the DB.
  During the loading of data, the application will create work tables and then remove it when the process is complete. This recommended by Oracle on this application is to create a tablespace separated with no record.
  
  Here's the recommendation:
  "To minimize the contention of disk and logging, you can create a tablespace with NoLogging for the.
  work tables and indexes and store data files for storage on a physical disk separate.
  Because work tables are created and dropped during the processing of data, creating a tablespace
  without connect to work tables and the indexes of the working table can improve performance. »
  
  I asked experts to demand that all agree to implement the above suggestion. However, our DBA is reluctant to implement the suggestion above that there are concerns about the validity of the hot backup. Here is his comment:
  "The problem is that we cannot guarantee the recovery of this database unless recover us from a cold backup. This means that if there is something on this database which is important, we will have problems with data protection. "He said, not to apply any log will affect the ability to recover the database in the event of failure, even if tables work East of temporary tables only. His opinion is that, when the backup occur at the same time when the table is created, will result not valid backup. The only way to save the database must publish a cold backup, which means that we must stop the application to make the backup.
  
  Is it true. What is the best backup strategy for the above case?
  From the expert point of view of the application, cold backup is not necessary that the application does not maintain a persistent connection to the database.
  
  Thank you.

  If it's a temporary table only that is kept in storage space, in any case the data on their part would go, either at the end of the transaction or at the end of the session. Not sure that the data your dba trying to protect? I have yet to have my first cup of coffee then maybe missing me something obvious here, but I don't see a point in the protection of temporary tables and "cold backup to use for their recovery.

  HTH
  Aman...