Import of schemas and tables of RMAN backup

Similar Questions

• Cannot retrieve table of rman backup

  For the test, I want to recover the deleted table of rman backup.

  I do oracle 11g VM.  the next step, I've applied but could not recover the table.

  RMAN > backup database;

  On sqlplus create connect scott/Tiger

  create table scott.emp1 (col1 varchar2 (10));

  Insert into the value emp1 (col1) ('Young');

  commit;

  Select * from emp1.

  He shows me a single line successfully.

  Now, I have connection to the sys as sysdba

  sqlplus > alter system switch logfile;

  After I remove the table.

  Now to retrieve the table I restored the database on the same virtual machine.

  RMAN > restore database;

  RMAN > restore database;

  RMAN > alter database open;

  sqlplus I've issued

  sqlplus > select * from scott.emp1;

  table or view does not exist.

  any help how to recover the deleted table of rman backup

  Hello

  Check this test scenario

  SQL > create table TEST2 in select * from emp;

  Table created.

  SQL > select count (*) from TEST2;

  COUNT (*)

  ----------

  15

  SQL > select TO_CHAR (SYSDATE, "MON-DD-YYYY HH24:MI:SS) twice;

  TO_CHAR (SYSDATE,'DD-)

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

  2013-SEPT-10 10:57:05

  SQL > alter system switch logfile;

  Modified system.

  SQL > drop table TEST2 is serving;

  Deleted table.

  SQL > immediate shu

  rman target $ [oracle@vnode MYSHELL].

  Recovery Manager: release 11.2.0.2.0 - Production kills Sep 10 10:58:39 2013

  Copyright (c) 1982, 2009, Oracle and/or its affiliates.  All rights reserved.

  connected to target database: DDTEST (DBID = 1667776968, is not open)

  RMAN > run {}

  until ' to_date (' 10 10:57 ',' yyyy Sep 2013 my dd HH24 ') ";

  restore the database;

  recover the database;

  ALTER database open resetlogs;

  } 2 > 3 > 4 > 5 > 6 >

  ...........

  ...........

  media recovery complete, duration: 00:00:02

  Finished recover 13 / SEP / 10

  open database

  RMAN > exit

  SQL > define pages of lin 200 100

  SQL > select count (*) from test2;

  COUNT (*)

  ----------

  15

  HTH

• CC of Lightroom does not import the photos and displays an error, backup also fails

  Hello

  I use Lightroom CC on Mac and I try to import images, but Lightroom is unable to do it this time for some reason any.

  During the last import Lightroom does not have the download of pictures and showed me the following notification: "Lightroom has encountered an error when reading the preview cache and needs to stop smoking. Lightroom will attempt to solve this problem, the next time it starts.

  As it stops it suggests to the backup catalog files and the following notification appears: "Lightroom could not save the catalog named XXX.". Check your folder permissions, and then make sure you have space on your rear disc and disc from the main catalogue.

  I have 2 TB of disk space available.

  I tried to import the camera and my computer, however it is unable to complete the import in two ways, just that several files have been imported and import fails. In each attempt to import files Lightroom creates temporary files with multidigital names.

  
  

  I restarted Lightroom, then my computer, then attempted to create another backup folder, nothing happens. Please help me.

  Thank you.

  Hi all

  6.3/CC 2015.3 Lightroom is now available - try to update to this version which includes the previous import feature, bug fixes and new support for photo/target device profile: Lightroom CC 2015.3 / 6.3 now available

  Concerning

  Pete

• What is the diff between backupset and element of RMAN backup

  Hello
  
  Please help me understand the diff backupset with RMAN backup element
  
  Thank you

  Hello

  Are you can, use the BACKUP AS BACKUPSET command, you can have a multiple backupsets.

  Kind regards
  Delphine K

• How to recover a table without rman backup

  Hello
  
  Database version: 10.2.0.4
  Operating system: windows 2003 r2
  
  I lost one table in my database of development, the size is 1 TB. Will have only a manual backup.
  I want to restore the table, not the full base because the size is 1 TB.
  
  We do not use rman and flashback is not enabled.
  
  Thank you and best regards,

  user12133150 wrote:
  Hello

  Table was droped. Recyclebin is disabled in my database.

  Y at - it another option. I have not also export dump.

  Thank you

  No.You must restore your database on server/members and open this database. Take then export this table and finally that import into the production database

• Car cloning + asm_cold RMAN backup

  Hello
  
  I have a db v11.2 RAC, that uses asm and saved using rman (backup, back up to asm diskgroup d cold).
  
  is there any document on how to clone this db on the same cluster or remote cluster?
  If I clone on the same cluster so I guess I should give it a new name too...
  
  
  pouvez cloning change db characterset as well?
  
  Edited by: user9198889 January 6, 2011 08:49

  Hopes of the documents below will help,

  Need to Document/steps of cloning RAC 11 g.
  http://download.Oracle.com/docs/CD/E14072_01/RAC.112/e10718/clonerac.htm
  http://blogs.Oracle.com/AlejandroVargas/gems/RMANDUPDBPRACTICE.PDF

  Thank you

• Tables of export and import the schema using the system user

  Hello
  
  Can someone asssit me please:
  
  I want to export all the tables / or some tables of a schema (ETLUSER) on DB1 to ETLUSER on an other (DB2) database by using the user of the SYSTEM. The SYSTEM user has DBA privilege.
  How I do what I do not know the password of ETLUSER.
  
  Environment:
  Oracle 11 g 2, AIX5
  
  Thank you.

  no problem, shouldn't you password in the ETLUSER.
  Start your exp or imp of the user of the system

  below, scott table dept exported and imported in the HR schema.

  Oracle@oragg1 > exp tables = scott.dept file = a.dmp log = b.log

  Export: Release 10.2.0.4.0 - Production on Mon Mar 19 18:58:45 2012

  Copyright (c) 1982, 2007, Oracle. All rights reserved.

  Username: System
  Password:

  Connected to: Oracle Database 10g Enterprise Edition Release 10.2.0.4.0 - Production 64-bit
  With the partitioning, Real Application Clusters, OLAP, data mining
  and Real Application Testing options
  Export in US7ASCII and AL16UTF16 NCHAR character set
  Server uses WE8ISO8859P1 (possible character set conversion) character set

  About to export specified tables by conventional means...
  Current user has changed to SCOTT
  . . export the DEPT 10 exported table rows

  Oracle@oragg1 > imp leader = a.dmp log = bb.log fromuser = scott touser = hr

  Import: Release 10.2.0.4.0 - Production on Mon Mar 19 18:58:56 2012

  Copyright (c) 1982, 2007, Oracle. All rights reserved.

  Username: System
  Password:

  Connected to: Oracle Database 10g Enterprise Edition Release 10.2.0.4.0 - Production 64-bit
  With the partitioning, Real Application Clusters, OLAP, data mining
  and Real Application Testing options

  Export file created by EXPORT: V10.02.01 by conventional means
  import in US7ASCII and AL16UTF16 NCHAR character set
  import server uses WE8ISO8859P1 (possible character set conversion) character set
  *. import of SCOTT in HR objects *. .
  "DEPT" table import 10 lines imported
  Import completed successfully without warnings.

• Transformation to define the schema and user on Tables in relational and physical

  I imported my Designer SDDM 4.0.1.836 deposit.  Now, I want to APP_SYS of subviews allows to define the schema on the tables in the relational model and the user on the physical model.  Our application in designer systems are 90% based scheme.  Here's what I have so far based on the example of OE and HR plans that I imported, built logical, then back to relational.  At this point, I have the tables by subview...  Line 24 is obviously incorrect.  Can someone give me a step by step approach to understand what I need to set the schema of the table and the user?  I looked through the read me file, xml files, the index and searched the known universe and am still confused.

  importPackage(javax.swing);
  // variable to keep a status message for later
  
  var message="";
  var schema="";
  
  // Get all subviews for Relational
      subviews = model.getListOfSubviews().toArray();
  for (var isub = 0; isub<subviews.length;isub++){
      subview = subviews[isub];
      message = message + '\n'+isub+" "+ subviews[isub];
      if(subviews[isub]!=null){
            if(subview == "HR_ERD"){
               shema = "HR";
              }else if(subview == "OE_ERD"){
               schema = "OE";
              }
           tables = model.getTableSet().toArray();
           for (var t = 0; t<tables.length;t++){
               table = tables[t];
               //check for presentation in subview
               tv = table.getFirstViewForDPV(subviews[isub]);
               if(tv!=null){
                   table.setSchemaObject(schema);  
                    message = message + '\n    '+ t +" " +shema+"."+ tables[t] ;
              }
          }
      }
  }  
  
  JOptionPane.showMessageDialog(null, message);
  //Packages.oracle.dbtools.crest.swingui.ApplicationView.log(message);
  

  Hi Marcus,

  First of all, you must make sure that the schema already exists in your relational model.

  You can use the method

  public void setSchema (String schemaName)

  to associate your Table with this scheme.

  (This method is defined on ObjetConteneur, so it can also be used to make and index).

  If you already have a user of your physical model that implements your schema, it is not normally necessary to create a user on the TableProxyOracle in the physical model.

  However if you do not want to set, then the method to use is

  public void setUser (user UserOracle)

  David

• Implications of losing the directory ArchivedLogs and RMAN backup.

  DB version: 10 gr 2
  We have a database of test that is important, since this is a pre-production DB used by the application for the final test team.
  
  Currently this DB is not in Archivelog mode. Every night a cron job runs to create the dump of datapump expdp of patterns in this PB.
  
  We want to test the RMAN backups online in this PB. So, let's put this db in ARCHIVELOG mode for a week.
  
  The RMAN backup items will be stored in c++ / u03/data1/orabkp + and archived logs will be stored in c++ / u04 / directory ach_logs +.
  
  We want to know the impact of/u04/ach_logs, the directory of archived newspapers getting accidentally deleted. I know that the accidental removal of the RMAN backup directory should not affect the normal operation of the database. But we want to know if the accidental deletion of archive redo logs directory will affect the DB.
  
  Published by: user10633418 on 25 Sep, 2009 02:53

  NO, it will not affect the database. If you delete the files of the database archive logs will work normally. But you restore your database to the point in time. For recovery, you need your full backup of all data files + archivelogs online redo log files.

  For a policy of backup of a production database where you won't lose any transaction that you must book your archive saves files of rman backup.

  Concerning
  Asif Kabir

• Merits and demerits of the replication of schema-level and Table

  Hello
  
  We evaluate the streams for our needs of replication. We reproduce nearly 2000 two-way tables. Most of the tables are in a schema. The main schema has a few tables that are not replicated.
  
  Replicate us schema level? Or at the level of the table? What are the advantages and disadvantages?
  
  Replication of level schema included with a handful of exclusionary rules seems simpler.
  
  In water courses at the level of the tables as well as creating 2000 table rules in the beginning, add a new table to an existing configuration of streams, will require additional steps to stop capture/apply/spread, creating new rules for the new table and restart replication. However, it seems that level replication table offering flexibility in terms of group tables. We wanted to group tables about 10 groups of 20 to 300 tables in each group. Each group will have its own capture, propagate and apply. In this way, we can isolate large groups of those unimportant?
  
  Who is a better setup in terms of maneuverability, flexibility, parallelism and performance?
  
  Thank you.

  Hello Stargaze,

  OK, so I understand that the grouping is based on the type of transactions and its importance. In this case the grouping would help you in many ways such as:

  1. If you encounter a problem or a bug with one of the processes of capture/apply then the other groups would work fine. Therefore, you have the opportunity to shoot a game of capture, propagation and applies until solve you the problem.

  2. in case if you perform any load batch where the transaction will affect over 10000 + records then it is possible that you could get stuck at 'SUSPENDED for FLOW CONTROL' on any of the feed items. With different groups would give you more flexibility to solve this problem by activating and deactivating groups.

  For more information on this please check the following:

  Apply process is active, inactive, and just stopped in change application

  3. with the schema level replication even if you have the advantage of being simple and have fewer rules in the set of rules, you must consider the 2 points above before you implement it. You need to test both diagrams and tables with a typical load (simulation of the production) and need to identify potential problems in both. Some more points to consider:

  -apply the game patches 10.2.0.4 or at least 10.2.0.3
  -apply all the patches required streams (Note: 437838.1> on metalink)
  -follow the recommendations as indicated:
       Note: 418755.1 broadcast 10.2.0.x.x of recommendations
       Note: 335516.1 Performance of flow recommendations
  Note: 413353.1 10.2 best practices for streams in RAC environment

  Thank you
  Florent

• RMAN backup location must change and remove obsolete

  Dear Experts,

  I'm working on the following environment,

  Operating system: Windows server 2012 R2

  Oracle version: 11.2.0.1.0 release

  Type of RMAN backup: Cumulative (preferred by the direction and I can't change it in differential backup)

  My current situation,

  I set up RMAN and H: drive for purposes of backup and G: for Archivelog.

  Recently, my city was covered by the flood waters and the server works not for a week so that we delete obsolete level 0 does not work.

  Currently there is not enough space to run a level 1 and level 0, while I should change the backup location.

  My Question is,

  (1) I need to change this location of drive H: to G: drive backup backup correctly without error not enough space.

  (2) after having taken a level 1 and level 0, depending on my 1 redundancy, RMAN mechanism should delete obsolete backup disk without failure G: and H:.

  Kindly help me how to get to my need. This will help me, thanks in advance

  You can

  1. replace the DB_RECOVERY_FILE_DEST G:

  2 run backups

  3. wait for the DELETE OBSOLETE delete obsolete backups of H:

  and then

  4. change the DB_RECOVERY_FILE_DEST return h as it was more early

  In this way, you don't have to use the FORMAT or CHANNEL of CONFIGURE commands (I was referring to CONFIGURE CHANNEL, not ALLOCATE CHANNEL) to change the destination of the backup

  Hemant K Collette

• content managed rman backup crushing and selection procedure of the tape media

  Hello Experts.

  I have a question about the backupset rman tape media manage tape of the osb.

  [ENV]

  Exadata rman backup assistance OSB(version: 10.4.0.3.0) SL500 LTO5 4 Tape Drive with rman scripts and full backup is working well.

  1 Appendix backup on OS crontab

  -Each backup full day at 00:00 AM

  2 rman view all:

  ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

  using the control file of the target instead of recovery catalog database

  RMAN settings for database with db_unique_name TESTDW are:

  CONFIGURE RETENTION POLICY TO 3 DAY RECOVERY WINDOW;

  CONFIGURE BACKUP OPTIMIZATION # by default

  CONFIGURE THE TYPE OF DEFAULT DEVICE TO 'SBT_TAPE ';

  SET UP THE DEVICE TYPE 'SBT_TAPE' PARALLELISM 4 TYPE OF BACKUP BACKUPSET.

  CONFIGURE CONTROLFILE AUTOBACKUP ON;

  ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

  3 rman backup scirpts

  Run {}

  allocate channels c01 type 'sbt_tape' connect 'sys/manager@testdw1';

  allocate channel c02 type 'sbt_tape' connect 'sys/manager@testdw2';

  allocate channel c03 type 'sbt_tape' connect 'sys/manager@testdw1';

  allocate channels c04 type 'sbt_tape' connect 'sys/manager@testdw2';

  Backup section size 900g database;

  output channel c01;

  output channel c02;

  output channel c03;

  output channel c04;

  }

  ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

  1st quarter > after several weeks and I run the command as "#obtool u admin Pei welcome1 lsvol-L lib1.

  and are used in many media of rman backup band (about 50EA media, perhaps 46EA are full and 4EA are appendble)

  So, I issue the command 'lsvol - a - c' and could not display expired on all media content.

  Why OSB no has not select and replace the tape media despite the rman rman default retention policy (3days) all show the setting.

  Why OSB does not have the expired status of rman content on tape media?

  -> I need the rman (delete expired, delete obsolete etc...) command to share the summary status of OSB and rman rman?

  Q2 > selection procedure for the tape media all backup rman

  Suppose the total number of tape media is 5EA. (1,2,3,4,5) with an available tape LTO5 drive.

  Status: 1 (full-rman content already expired), 2 (content integral-rman not expired), 3 (editable, remaining space), 4 (new media), 5 (currnet wring media)

  If OSB burn now 5 and reach at the end of the Strip and possibly another media, media to be selected by the OSB?

  Advanced thanks.

  If I place the order obsolete in the rman script as 'delete type 'SBT_TAPE' noprompt obsolete device;' the results will be the same as your

  the syntax because rman default setting (CONFIGURE RETENTION POLICY to RECOVERY WINDOW OF 3 DAYS ;) has already been fixed on the command "Show all", right? @)

  Yes.

  I use the catalog of recovery on a different database and backup script has catalog db connection string.

  And if I only use controlfile instead of recovery catalog to save the backup information and put in place "control_file_record_keep_time" a default of 7 days to 365 days, "remove obsolete" command to remove several weeks backupset works well, is it not?

  Possible problem are existing backups for several weeks if you left default control_file_record_keep_time. Change now will not help unless you catalog later manually all the old pieces not catalogued on tape backup.

  After you clean up the situation and particularly to remove obsolete in your script, rman, you will never make problems with control_file_record_keep_time from this moment. By default will do just fine because it is larger than your window of recovery.

  Kind regards

  Mitja

• RMAN backup and restore into Production KEEPS RAC and DATA

  Hi gurus of the Oracle
  
  We need the expert advice for tune and cross check the RMAN Script for Production environment below below
  
  Please suggest any corrections or comments on below mention scripts which will be executed on RAC and Production DATA GUARD environment
  
  LINUX RHEL 5.5 64-bit
  ORACLE GRID Infrastructure R2 Version 11.2.0.3.0 (we use ASM Cluster File system)
  ORACLE11g R2 Version 11.2.0.3.0
  
  We planned for basically rman backup drive using control file target instead of recovery catalog database
  
  We will have Data Guard Setup between 2 remote sites of database and two remote sites have running the local 2-node RAC database
  
  =========================RMAN_Full_Backup.sh=========================================================================
  #! / bin/bash
  . /Home/orarac/.bash_profile
  logloc = 'date' + rman_backup_lvl0_orcldb_ %m %y hour %M %S %d ".log
  # Start rman backup
  Target of $ORACLE_HOME/bin/rman = / log = / backup/RMAN_Log / $logloc < < EOF
  run
  {
  allocate channels ch1 type disk maxpiecesize = 1024 M;
  allocate channel ch2 type disk maxpiecesize = 1024M;
  database backup incremental level 0 FORMAT ' / backup/RMAN_Backup/rman_bkp_%U_%d_%T_DATAFILE_LVL0_%D_%M_%Y'
  include current controlfile
  tag 'datafile_full ';
  }
  
  run
  {
  allocate channels ch1 type disk maxpiecesize = 1024 M;
  allocate channel ch2 type disk maxpiecesize = 1024M;
  overlap archivelog all;
  SQL 'alter system archive log current';
  backup ARCHIVELOG all delete them all input FORMAT ' / backup/RMAN_Backup/rman_bkp_%U_%d_%T_ARCHIVELOG_%D_%M_%Y'
  tag 'archivelog_full ';
  }
  cross-checking of backup;
  delete noprompt obsolete;
  Save the list of database;
  EXIT;
  EXPRESSIONS OF FOLKLORE
  ==================================================================================================
  =========================================RMAN_Inc_Backup.sh ===========================================
  #! / bin/bash
  . /Home/orarac/.bash_profile
  logloc = 'date' + rman_backup_lvl1_orcldb_ %m %y hour %M %S %d ".log
  # Start rman backup
  # use only the following line if this script is run via the root CRON
  Target of $ORACLE_HOME/bin/rman = / log = / backup/RMAN_Log / $logloc < < EOF
  
  Run {}
  allocate channels ch1 type disk maxpiecesize = 1024 M;
  allocate channel ch2 type disk maxpiecesize = 1024M;
  database backup incremental level 1 FORMAT ' / backup/RMAN_Backup/rman_bkp_%U_%d_%T_DATAFILE_LVL1_%D_%M_%Y'
  include current controlfile
  tag 'datafile_Incr ';
  }
  Run {}
  allocate channels ch1 type disk maxpiecesize = 1024 M;
  allocate channel ch2 type disk maxpiecesize = 1024M;
  overlap archivelog all;
  SQL 'alter system archive log current';
  backup ARCHIVELOG all delete them all input FORMAT ' / backup/RMAN_Backup/rman_bkp_%U_%d_%T_ARCHIVELOG_%D_%M_%Y'
  tag 'archivelog_full ';
  }
  cross-checking of backup;
  delete noprompt obsolete;
  backup from the list;
  EXIT;
  EXPRESSIONS OF FOLKLORE
  -bash - $3.2
  ====================================================================================================
  ===================================RMAN_DB_Restore.sh====================================================
  -bash-3, $ 2 cat /home/orarac/RMAN_Scripts/RMAN_DB_Restore.sh
  #! / bin/bash
  . /Home/orarac/.bash_profile
  logloc = 'date' + rman_restore_orcldb_ %m %y hour %M %S %d ".log
  # use only the following line if this script is run via the root CRON
  Target of $ORACLE_HOME/bin/rman = / log = / backup/RMAN_Log / $logloc < < EOF
  Run {}
  BOOTABLE MEDIA FORCE;
  RESTORE THE DATABASE;
  RECOVER THE DATABASE;
  ALTER DATABASE OPEN;
  }
  EXIT;
  EXPRESSIONS OF FOLKLORE
  ====================================================================================================
  
  Concerning
  Hitgon

  Thank you mseberg

  This command is the full backup or a backup of level 0?

  full backup as compressed backupset
  data include current controlfile more archivelog deletion entry;

  We use command below for the full backup of database.
  database backup incremental level 0 FORMAT ' / backup/RMAN_Backup/rman_bkp_%U_%d_%T_DATAFILE_LVL0_%D_%M_%Y'
  include current controlfile
  tag 'datafile_full ';

  We use the command below to the increment backup
  database backup incremental level 1 FORMAT ' / backup/RMAN_Backup/rman_bkp_%U_%d_%T_DATAFILE_LVL1_%D_%M_%Y'
  include current controlfile
  tag 'datafile_Incr ';

  What is the different when we use
  CONFIGURE ARCHIVELOG DELETION POLICY TO APPLIED ON STANDBY;

  or

  full backup as backupset compressed database include controlfile current more archivelog delete all input;

  Concerning
  Hitgon

• CREATE TABLE to another OWNER/SCHEMA and in a different TABLESPACE?

  I am connected to a user of the SYSTEM. Now, I want to create a table of aaa. The owner of this table should not be SYSTEM but user KARL. and STORAGE space should not be SYSTEM but the tdtc TABLESPACE (existing).
  
  As far as I know I can do this by running the following command:
  
  CREATE TABLE KARL.aaa (an INTEGER,...) Tttt TABLESPACE;
  
  Regarding the TABLESPACE parameter, I'm not sure. Is it possible to return a user Karl TABLE in a TABLESPACE that is not assigned to him?
  
  In addition, I have another problem.
  I have a script with hundreds of CREATE TABLE, ALTER TABLE + CREATE INDEX DDL statements.
  Each of them are not prefixed with schema/owner and a TABLESPACE clause.
  
  I can put a single statement at the top of the script somehow that such Oracle
  to use
  -Karl OWNER as a schema/owner for all DDL stements
  -Thomas TABLESPACE as TABLESPACE for all DOF stements
  ?
  
  In MYsql, there is a statement «use < database >» Is soemthing similar to Oracle?
  
  Thank you
  Peter

  Yes... you can do. For example, a user who has a secret password that you do not want to give... as a schema of the application. User B needs to make paintings/objects in A schema and you want to follow this as user B. First configure Oracles Fine grain auditing and then grant "connect via" user A to user B as follows:
  -----

  SQL> create user b identified by abc123
    2    quota unlimited on users;
  
  User created.
  SQL> grant create session to b;
  
  SQL> create user a identified by abc123
    2    quota unlimited on users;
  
  User created.
  
  SQL> grant create table
    2      , create session
    3     to a;
  
  Grant succeeded.
  
  SQL> alter user a grant connect through b;
  
  User altered.
  
  SQL> connect b[a]/abc123@a486
  Connected.
  
  SQL> show user
  USER is "A"
  SQL> create table a.my_proxy_table
    2  ( c1 number
    3  , c2 varchar2(50)
    4  , c3 date
    5  );
  
  Table created.
  

  -----
  See I have connected using the syntax "username [proxyuser]." Also note that user B is by proxy user has, as evidenced by my order 'show user '. The connection through can be granted/revoked as needed without disclosing A goes to B. In addition, your audit tables or xml logs will follow the fact that user B created the table a.my_proxy_table.

  Hope this helps,
  John

• Why import data if ready schema and index create, increase in Undo TableSpace

  I have a problem importing data.
  When I export data don't have schema import include the index schema.
  And then I import the schema ready.
  When I import data, I find this Undo TableSpace increase very fast.
  I have therefore two tests.
  1. IMPORT of data and index, every thing ok look
  2 import a schema includes index
  and then import data. Undo Tablespace very rapid increase

  If you mean undo tablespace increase in order to read consisten read.

  N ° cancel space use increases due to the DML. (requests for consistent readings read undo, they generate no cancellation to grow)

  so you're saying that redo log appear, rollback undo segment will increase.

  N ° increase in written cancellations results increase in roll forward.

  IF we import large data, we must export data with the index and import data right now

  If you import large data, you must create the indexes after the data is imported. If the index definitions are included in the export, the default behavior is that import will create them once the data is imported. To do this, undo lower for indexes.
  Note This will generate redo for the creation (but does not cancel and therefore not generate recovery for cancellation). You can avoid this by manually running CREATE INDEX... NOLOGGING - but it means you need to get and modify the scripts CREATE INDEX.

  Hemant K Collette