Skip Headers
Oracle® Database Backup and Recovery User's Guide
11g Release 1 (11.1)

Part Number B28270-01
Go to Documentation Home
Home
Go to Book List
Book List
Go to Table of Contents
Contents
Go to Index
Index
Go to Master Index
Master Index
Go to Feedback page
Contact Us

Go to previous page
Previous
Go to next page
Next
View PDF

27 Making User-Managed Database Backups

This chapter describes methods of backing up an Oracle database in a user-managed backup and recovery strategy, that is, a strategy that does not depend on using Recovery Manager (RMAN).

This chapter contains the following sections:

Querying V$ Views to Obtain Backup Information

Before making a backup, you must identify all the files in your database and decide what to back up. You can use V$ views to obtain this information.

Listing Database Files Before a Backup

Use V$DATAFILE and V$CONTROLFILE to identify the datafiles and control files for your database. This same procedure works whether you named these files manually or allowed Oracle Managed Files to name them.

Caution:

Never back up online redo log files.

To list datafiles and control files:

  1. Start SQL*Plus and query V$DATAFILE to obtain a list of datafiles. For example, enter:

    SELECT NAME FROM V$DATAFILE;
    

    You can also join the V$TABLESPACE and V$DATAFILE views to obtain a listing of datafiles along with their associated tablespaces:

    SELECT   t.NAME "Tablespace", f.NAME "Datafile"
    FROM     V$TABLESPACE t, V$DATAFILE f
    WHERE    t.TS# = f.TS#
    ORDER BY t.NAME;
    
  2. Obtain the filenames of the current control files by querying the V$CONTROLFILE view. For example, issue the following query:

    SELECT NAME FROM V$CONTROLFILE;
    

    Note that you only need to back up one copy of a multiplexed control file.

  3. If you plan to take a control file backup with the ALTER DATABASE BACKUP CONTROLFILE TO 'filename' statement, then save a list of all datafiles and online redo log files with the control file backup. Because the current database structure may not match the database structure at the time a given control file backup was created, saving a list of files recorded in the backup control file can aid the recovery procedure.

Determining Datafile Status for Online Tablespace Backups

To check whether a datafile is part of a current online tablespace backup, query the V$BACKUP view.

This view is useful only for user-managed online tablespace backups, because neither RMAN backups nor offline tablespace backups require the datafiles of a tablespace to be in backup mode.

The V$BACKUP view is most useful when the database is open. It is also useful immediately after an instance failure because it shows the backup status of the files at the time of the failure. Use this information to determine whether you have left any tablespaces in backup mode.

V$BACKUP is not useful if the control file currently in use is a restored backup or a new control file created after the media failure occurred. A restored or re-created control file does not contain the information the database needs to populate V$BACKUP accurately. Also, if you have restored a backup of a file, this file's STATUS in V$BACKUP reflects the backup status of the older version of the file, not the most current version. Thus, this view can contain misleading data about restored files.

For example, the following query displays which datafiles are currently included in a tablespace that has been placed in backup mode:

SELECT t.name AS "TB_NAME", d.file# as "DF#", d.name AS "DF_NAME", b.status
FROM   V$DATAFILE d, V$TABLESPACE t, V$BACKUP b
WHERE  d.TS#=t.TS#
AND    b.FILE#=d.FILE#
AND    b.STATUS='ACTIVE';

The following sample output shows that the tools and users tablespaces currently have ACTIVE status:

TB_NAME                 DF#        DF_NAME                           STATUS
----------------------  ---------- --------------------------------  ------
TOOLS                   7          /oracle/oradata/trgt/tools01.dbf  ACTIVE
USERS                   8          /oracle/oradata/trgt/users01.dbf  ACTIVE

In the STATUS column, NOT ACTIVE indicates that the file is not currently in backup mode (that is, you have not executed the ALTER TABLESPACE ... BEGIN BACKUP or ALTER DATABASE BEGIN BACKUP statement), whereas ACTIVE indicates that the file is currently in backup mode.

Making User-Managed Backups of the Whole Database

You can make a whole database backup of all files in a database after the database has been shut down with the NORMAL, IMMEDIATE, or TRANSACTIONAL options. A whole database backup taken while the database is open or after an instance failure or SHUTDOWN ABORT is inconsistent. In such cases, the files are inconsistent with respect to the checkpoint SCN.

You can make a whole database backup if a database is operating in either ARCHIVELOG or NOARCHIVELOG mode. If you run the database in NOARCHIVELOG mode, however, then the backup must be consistent; that is, you must shut down the database cleanly before the backup.

The set of backup files that results from a consistent whole database backup is consistent because all files are checkpointed to the same SCN. You can restore the consistent database backup without further recovery. After restoring the backup files, you can perform additional recovery steps to recover the database to a more current time if the database is operated in ARCHIVELOG mode. Also, you can take inconsistent whole database backups if your database is in ARCHIVELOG mode.

Control files play a crucial role in database restore and recovery. For databases running in ARCHIVELOG mode, Oracle recommends that you back up control files with the ALTER DATABASE BACKUP CONTROLFILE TO 'filename' statement.

See Also:

"Making User-Managed Backups of the Control File" for more information about backing up control files

Making Consistent Whole Database Backups

This section describes how to back up the database with an operating system utility.

To make a consistent whole database backup:

  1. If the database is open, then use SQL*Plus to shut down the database with the NORMAL, IMMEDIATE, or TRANSACTIONAL options.

  2. Use an operating system utility to make backups of all datafiles as well as all control files specified by the CONTROL_FILES parameter of the initialization parameter file. Also, back up the initialization parameter file and other Oracle product initialization files. To find these files, do a search for *.ora starting in your Oracle home directory and recursively search all of its subdirectories.

    For example, you can back up the datafiles, control files and archived logs to /disk2/backup as follows:

    % cp $ORACLE_HOME/oradata/trgt/*.dbf /disk2/backup
    % cp $ORACLE_HOME/oradata/trgt/arch/* /disk2/backup/arch
     
    
  3. Restart the database with the STARTUP command in SQL*Plus.

    See Also:

    Oracle Database Administrator's Guide for more information on starting up and shutting down a database

Making User-Managed Backups of Tablespaces and Datafiles

The technique for making user-managed backups of tablespaces and datafiles depends on whether the files are offline or online.

Making User-Managed Backups of Offline Tablespaces and Datafiles

Note the following guidelines when backing up offline tablespaces:

  • You cannot offline the SYSTEM tablespace or a tablespace with active undo segments. The following technique cannot be used for such tablespaces.

  • Assume that a table is in tablespace Primary and its index is in tablespace Index. Taking tablespace Index offline while leaving tablespace Primary online can cause errors when DML is issued against the indexed tables located in Primary. The problem only manifests when the access method chosen by the optimizer needs to access the indexes in the Index tablespace.

To back up offline tablespaces:

  1. Before beginning a backup of a tablespace, identify the tablespace's datafiles by querying the DBA_DATA_FILES view. For example, assume that you want to back up the users tablespace. Enter the following statement in SQL*Plus:

    SELECT TABLESPACE_NAME, FILE_NAME
      FROM SYS.DBA_DATA_FILES
      WHERE TABLESPACE_NAME = 'USERS';
     
    TABLESPACE_NAME                   FILE_NAME
    -------------------------------   --------------------------------
    USERS                             /oracle/oradata/trgt/users01.dbf
    

    In this example, /oracle/oradata/trgt/users01.dbf is a fully specified filename corresponding to the datafile in the users tablespace.

  2. Take the tablespace offline using normal priority if possible because it guarantees that you can subsequently bring the tablespace online without having to recover it. For example:

    SQL> ALTER TABLESPACE users OFFLINE NORMAL;
    
  3. Back up the offline datafiles. For example:

    % cp /oracle/oradata/trgt/users01.dbf /d2/users01_'date "+%m_%d_%y"'.dbf
    
  4. Bring the tablespace online. For example:

    ALTER TABLESPACE users ONLINE;
    

    Note:

    If you took the tablespace offline using temporary or immediate priority, then you cannot bring the tablespace online unless you perform tablespace recovery.
  5. Archive the unarchived redo logs so that the redo required to recover the tablespace backup is archived. For example, enter:

    ALTER SYSTEM ARCHIVE LOG CURRENT;
    

Making User-Managed Backups of Online Tablespaces and Datafiles

You can back up all or only specific datafiles of an online tablespace while the database is open. The procedure differs depending on whether the online tablespace is read/write or read-only.

Note:

You should not back up temporary tablespaces.

Making User-Managed Backups of Online Read/Write Tablespaces

You must put a read/write tablespace in backup mode to make user-managed datafile backups when the tablespace is online and the database is open. The ALTER TABLESPACE ... BEGIN BACKUP statement places a tablespace in backup mode. In backup mode, the database copies whole changed data blocks into the redo stream. After you take the tablespace out of backup mode with the ALTER TABLESPACE ... END BACKUP or ALTER DATABASE END BACKUP statement, the database advances the datafile header to the current database checkpoint.

When restoring a datafile backed up in this way, the database asks for the appropriate set of redo log files to apply if recovery be needed. The redo logs contain all changes required to recover the datafiles and make them consistent.

To back up online read/write tablespaces in an open database:

  1. Before beginning a backup of a tablespace, identify all of the datafiles in the tablespace with the DBA_DATA_FILES data dictionary view. For example, assume that you want to back up the users tablespace. Enter the following:

    SELECT TABLESPACE_NAME, FILE_NAME
    FROM   SYS.DBA_DATA_FILES
    WHERE  TABLESPACE_NAME = 'USERS';
     
    TABLESPACE_NAME                   FILE_NAME
    -------------------------------   --------------------
    USERS                             /oracle/oradata/trgt/users01.dbf
    USERS                             /oracle/oradata/trgt/users02.dbf
    
  2. Mark the beginning of the online tablespace backup. For example, the following statement marks the start of an online backup for the tablespace users:

    SQL> ALTER TABLESPACE users BEGIN BACKUP;
    

    Caution:

    If you do not use BEGIN BACKUP to mark the beginning of an online tablespace backup and wait for this statement to complete before starting your copies of online tablespaces, then the datafile copies produced are not usable for subsequent recovery operations. Attempting to recover such a backup is risky and can return errors that result in inconsistent data. For example, the attempted recovery operation can issue a "fuzzy files" warning, and can lead to an inconsistent database that you cannot open.
  3. Back up the online datafiles of the online tablespace with operating system commands. For example, Linux and UNIX users might enter:

    % cp /oracle/oradata/trgt/users01.dbf /d2/users01_'date "+%m_%d_%y"'.dbf
    % cp /oracle/oradata/trgt/users02.dbf /d2/users02_'date "+%m_%d_%y"'.dbf
    
  4. After backing up the datafiles of the online tablespace, run the SQL statement ALTER TABLESPACE with the END BACKUP option. For example, the following statement ends the online backup of the tablespace users:

    SQL> ALTER TABLESPACE users END BACKUP;
    
  5. Archive the unarchived redo logs so that the redo required to recover the tablespace backup is archived. For example, enter:

    SQL> ALTER SYSTEM ARCHIVE LOG CURRENT;
    

    Caution:

    If you fail to take the tablespace out of backup mode, then Oracle Database continues to write copies of data blocks in this tablespace to the online redo logs, causing performance problems. Also, you receive an ORA-01149 error if you try to shut down the database with the tablespaces still in backup mode.

Making Multiple User-Managed Backups of Online Read/Write Tablespaces

When backing up several online tablespaces, you can back them up either serially or in parallel. Use either of the following procedures depending on your needs.

Backing Up Online Tablespaces in Parallel

You can simultaneously create datafile copies of multiple tablespaces requiring backups in backup mode. Note, however, that by putting all tablespaces in online mode at once, you can generate large redo logs if there is heavy update activity on the affected tablespaces, because the redo must contain a copy of each changed data block in each changed datafile. Be sure to consider the size of the likely redo before using the procedure outlined here.

To back up online tablespaces in parallel:

  1. Prepare all online tablespaces for backup by issuing all necessary ALTER TABLESPACE statements at once. For example, put tablespaces users, tools, and indx in backup mode as follows:

    SQL> ALTER TABLESPACE users BEGIN BACKUP;
    SQL> ALTER TABLESPACE tools BEGIN BACKUP;
    SQL> ALTER TABLESPACE indx BEGIN BACKUP;
    

    If you are backing up all tablespaces, you might want to use this command:

    SQL> ALTER DATABASE BEGIN BACKUP;
    
  2. Back up all files of the online tablespaces. For example, a Linux or UNIX user might back up datafiles with the *.dbf suffix as follows:

    % cp $ORACLE_HOME/oradata/trgt/*.dbf /disk2/backup/
    
  3. Take the tablespaces out of backup mode as in the following example:

    SQL> ALTER TABLESPACE users END BACKUP;
    SQL> ALTER TABLESPACE tools END BACKUP;
    SQL> ALTER TABLESPACE indx END BACKUP;
    

    Again, it you are handling all datafiles at once you can use the ALTER DATABASE command instead of ALTER TABLESPACE:

    SQL> ALTER DATABASE END BACKUP;
    
  4. Archive the online redo logs so that the redo required to recover the tablespace backups will be available for later media recovery. For example, enter:

    SQL> ALTER SYSTEM ARCHIVE LOG CURRENT;
    
Backing Up Online Tablespaces Serially

You can place all tablespaces requiring online backups in backup mode one at a time. Oracle recommends the serial backup option because it minimizes the time between ALTER TABLESPACE ... BEGIN/END BACKUP statements. During online backups, more redo information is generated for the tablespace because whole data blocks are copied into the redo log.

To back up online tablespaces serially:

  1. Prepare a tablespace for online backup. For example, to put tablespace users in backup mode enter the following:

    SQL> ALTER TABLESPACE users BEGIN BACKUP;
    

    In this case you probably do not want to use ALTER DATABASE BEGIN BACKUP to put all tablespaces in backup mode simultaneously, because of the unnecessary volume of redo log information generated for tablespaces in online mode.

  2. Back up the datafiles in the tablespace. For example, enter:

    % cp /oracle/oradata/trgt/users01.dbf /d2/users01_'date "+%m_%d_%y"'.dbf
    
  3. Take the tablespace out of backup mode. For example, enter:

    SQL> ALTER TABLESPACE users END BACKUP;
    
  4. Repeat this procedure for each remaining tablespace.

  5. Archive the unarchived redo logs so that the redo required to recover the tablespace backups is archived. For example, enter:

    SQL> ALTER SYSTEM ARCHIVE LOG CURRENT;
    

Ending a Backup After an Instance Failure or SHUTDOWN ABORT

The following situations can cause a tablespace backup to fail and be incomplete:

  • The backup completed, but you did not run the ALTER TABLESPACE ... END BACKUP statement.

  • An instance failure or SHUTDOWN ABORT interrupted the backup.

Whenever crash recovery is required, if a datafile is in backup mode when an attempt is made to open it, then the database will not open the database until either a recovery command is issued, or the datafile is taken out of backup mode.

For example, the database may display a message such as the following at startup:

ORA-01113: file 12 needs media recovery
ORA-01110: data file 12: '/oracle/dbs/tbs_41.f'

If the database indicates that the datafiles for multiple tablespaces require media recovery because you forgot to end the online backups for these tablespaces, then so long as the database is mounted, running the ALTER DATABASE END BACKUP statement takes all the datafiles out of backup mode simultaneously.

In high availability situations, and in situations when no DBA is monitoring the database, the requirement for user intervention is intolerable. Hence, you can write a crash recovery script that does the following:

  1. Mounts the database

  2. Runs the ALTER DATABASE END BACKUP statement

  3. Runs ALTER DATABASE OPEN, allowing the system to come up automatically

An automated crash recovery script containing ALTER DATABASE END BACKUP is especially useful in the following situations:

  • All nodes in an Oracle Real Application Clusters (Oracle RAC) configuration fail.

  • One node fails in a cold failover cluster (that is, a cluster that is not a RAC configuration in which the secondary node must mount and recover the database when the first node fails).

Alternatively, you can take the following manual measures after the system fails with tablespaces in backup mode:

  • Recover the database and avoid issuing END BACKUP statements altogether.

  • Mount the database, then run ALTER TABLESPACE ... END BACKUP for each tablespace still in backup mode.

Ending Backup Mode with the ALTER DATABASE END BACKUP Statement

You can run the ALTER DATABASE END BACKUP statement when you have multiple tablespaces still in backup mode. The primary purpose of this command is to allow a crash recovery script to restart a failed system without DBA intervention. You can also perform the following procedure manually.

To take tablespaces out of backup mode simultaneously:

  1. Mount but do not open the database. For example, enter:

    SQL> STARTUP MOUNT
    
  2. If performing this procedure manually (that is, not as part of a crash recovery script), query the V$BACKUP view to list the datafiles of the tablespaces that were being backed up before the database was restarted:

    SQL>  SELECT * FROM V$BACKUP WHERE STATUS = 'ACTIVE';
    FILE#      STATUS             CHANGE#    TIME     
    ---------- ------------------ ---------- ---------
            12 ACTIVE                  20863 25-NOV-02
            13 ACTIVE                  20863 25-NOV-02
            20 ACTIVE                  20863 25-NOV-02
     3 rows selected.
    
  3. Issue the ALTER DATABASE END BACKUP statement to take all datafiles currently in backup mode out of backup mode. For example, enter:

    SQL> ALTER DATABASE END BACKUP;
    

    You can use this statement only when the database is mounted but not open. If the database is open, then use ALTER TABLESPACE ... END BACKUP or ALTER DATABASE DATAFILE ... END BACKUP for each affected tablespace or datafile.

    Caution:

    Do not use ALTER DATABASE END BACKUP if you have restored any of the affected files from a backup.
Ending Backup Mode with the SQL*Plus RECOVER Command

The ALTER DATABASE END BACKUP statement is not the only way to respond to a failed online backup: you can also run the SQL*Plus RECOVER command. This method is useful when you are not sure whether someone has restored a backup, because if someone has indeed restored a backup, then the RECOVER command brings the backup up to date. Only run the ALTER DATABASE END BACKUP or ALTER TABLESPACE ... END BACKUP statement if you are sure that the files are current.

Note:

The RECOVER command method is slow because the database must scan redo generated from the beginning of the online backup.

To take tablespaces out of backup mode with the RECOVER command:

  1. Mount the database. For example, enter:

    SQL> STARTUP MOUNT
    
  2. Recover the database as normal. For example, enter:

    SQL> RECOVER DATABASE
    
  3. Use the V$BACKUP view to confirm that there are no active datafiles:

    SQL>  SELECT * FROM V$BACKUP WHERE STATUS = 'ACTIVE';
    FILE#      STATUS             CHANGE#    TIME     
    ---------- ------------------ ---------- ---------
    0 rows selected.
    

See Also:

Chapter 28, "Performing User-Managed Database Flashback and Recovery" for information on recovering a database

Making User-Managed Backups of Read-Only Tablespaces

When backing up an online read-only tablespace, you can simply back up the online datafiles. You do not have to place the tablespace in backup mode because the database is not permitting changes to the datafiles.

If the set of read-only tablespaces is self-contained, then in addition to backing up the tablespaces with operating system commands, you can also export the tablespace metadata with the transportable tablespace functionality. In the event of a media error or a user error (such as accidentally dropping a table in the read-only tablespace), you can transport the tablespace back into the database.

See Also:

Oracle Database Administrator's Guide to learn how to transport tablespaces

To back up online read-only tablespaces in an open database:

  1. Query the DBA_TABLESPACES view to determine which tablespaces are read-only. For example, run this query:

    SELECT TABLESPACE_NAME, STATUS 
    FROM DBA_TABLESPACES
    WHERE STATUS = 'READ ONLY';
    
  2. Before beginning a backup of a read-only tablespace, identify all of the tablespace's datafiles by querying the DBA_DATA_FILES data dictionary view. For example, assume that you want to back up the history tablespace:

    SELECT TABLESPACE_NAME, FILE_NAME
    FROM SYS.DBA_DATA_FILES
    WHERE TABLESPACE_NAME = 'HISTORY';
     
    TABLESPACE_NAME                   FILE_NAME
    -------------------------------   --------------------
    HISTORY                           /oracle/oradata/trgt/history01.dbf
    HISTORY                           /oracle/oradata/trgt/history02.dbf
    
  3. Back up the online datafiles of the read-only tablespace with operating system commands. You do not have to take the tablespace offline or put the tablespace in backup mode because users are automatically prevented from making changes to the read-only tablespace. For example:

    % cp $ORACLE_HOME/oradata/trgt/history*.dbf  /disk2/backup/
    

    Note:

    When restoring a backup of a read-only tablespace, take the tablespace offline, restore the datafiles, then bring the tablespace online. A backup of a read-only tablespace is still usable if the read-only tablespace is made read/write after the backup, but the restored backup will require recovery.
  4. Optionally, export the metadata in the read-only tablespace. By using the transportable tablespace feature, you can quickly restore the datafiles and import the metadata in case of media failure or user error. For example, export the metadata for tablespace history as follows:

    % expdp DIRECTORY=dpump_dir1 DUMPFILE=hs.dmp TRANSPORT_TABLESPACES=history 
    > LOGFILE=tts.log
    

    See Also:

    Oracle Database Reference for more information about the DBA_DATA_FILES and DBA_TABLESPACES views

Making User-Managed Backups of the Control File

Back up the control file of a database after making a structural modification to a database operating in ARCHIVELOG mode. To back up a database's control file, you must have the ALTER DATABASE system privilege.

Backing Up the Control File to a Binary File

The primary method for backing up the control file is to use a SQL statement to generate a binary file. A binary backup is preferable to a trace file backup because it contains additional information such as the archived log history, offline range for read-only and offline tablespaces, and backup sets and copies (if you use RMAN). If COMPATIBLE is 10.2 or higher, binary control file backups include tempfile entries.

To back up the control file after a structural change:

  1. Make the desired change to the database. For example, you may create a new tablespace:

    CREATE TABLESPACE tbs_1 DATAFILE 'file_1.f' SIZE 10M;
    
  2. Back up the database's control file, specifying a filename for the output binary file. The following example backs up a control file to /disk1/backup/cf.bak:

    ALTER DATABASE BACKUP CONTROLFILE TO '/disk1/backup/cf.bak' REUSE;
    

    Specify REUSE to make the new control file overwrite one that currently exists.

Backing Up the Control File to a Trace File

You can back up the control file to a text file that contains a CREATE CONTROLFILE statement. You can edit the trace file to create a script that creates a new control file based on the control file that was current when you created the trace file.

If you specify neither the RESETLOGS nor NORESETLOGS option in the SQL statement, then the resulting trace file contains versions of the control file for both RESETLOGS and NORESETLOGS options. Tempfile entries are included in the output using ALTER TABLESPACE ... ADD TEMPFILE statements.

To avoid recovering offline normal or read-only tablespaces, edit them out of the CREATE CONTROLFILE statement. When you open the database with the re-created control file, the database marks these omitted files as MISSING. You can run an ALTER DATABASE RENAME FILE statement to rename them to their original filenames.

To back up the control file to a trace file:

  1. Mount or open the database.

  2. Execute the following SQL statement:

    ALTER DATABASE BACKUP CONTROLFILE TO TRACE;
    

    The trace files are stored in a subdirectory determined by the DIAGNOSTIC_DEST initialization parameter. To locate the directory for trace files, query the name and value columns of the V$DIAG_INFO dynamic performance view.

See Also:

"Recovery of Read-Only Files with a Re-Created Control File" for special issues relating to read-only, offline normal, and temporary files included in CREATE CONTROLFILE statements

Making User-Managed Backups of Archived Redo Logs

To save disk space in your primary archiving location, you may want to back up archived logs to tape or to an alternative disk location. If you archive to multiple locations, then only back up one copy of each log sequence number.

To back up archived redo logs:

  1. To determine which archived redo log files that the database has generated, query V$ARCHIVED_LOG. For example, run the following query:

    SELECT THREAD#,SEQUENCE#,NAME 
    FROM V$ARCHIVED_LOG;
    
  2. Back up one copy of each log sequence number by using an operating system utility. This example backs up all logs in the primary archiving location to a disk devoted to log backups:

    % cp $ORACLE_HOME/oracle/trgt/arch/* /disk2/backup/arch
    

    See Also:

    Oracle Database Reference for more information about the data dictionary views

Making User-Managed Backups in SUSPEND Mode

This section contains the following topics:

About the Suspend/Resume Feature

Some third-party tools allow you to mirror a set of disks or logical devices, that is, maintain an exact duplicate of the primary data in another location, and then split the mirror. Splitting the mirror involves separating the copies so that you can use them independently.

With the SUSPEND/RESUME functionality, you can suspend I/O to the database, then split the mirror and make a backup of the split mirror. By using this feature, which complements the backup mode functionality, you can suspend database I/Os so that no new I/O can be performed. You can then access the suspended database to make backups without I/O interference.

You do not need to use SUSPEND/RESUME to make split mirror backups in most cases, although it is necessary if your system requires the database cache to be free of dirty buffers before a volume can be split. Some RAID devices benefit from suspending writes while the split operation is occurring; your RAID vendor can advise you on whether your system would benefit from this feature.

The ALTER SYSTEM SUSPEND statement suspends the database by halting I/Os to datafile headers, datafiles, and control files. When the database is suspended, all pre-existing I/O operations can complete; however, any new database I/O access attempts are queued.

The ALTER SYSTEM SUSPEND and ALTER SYSTEM RESUME statements operate on the database and not just the instance. If the ALTER SYSTEM SUSPEND statement is entered on one system in a RAC configuration, then the internal locking mechanisms propagate the halt request across instances, thereby suspending I/O operations for all active instances in a given cluster.

Making Backups in a Suspended Database

After a successful database suspension, you can back up the database to disk or break the mirrors. Because suspending a database does not guarantee immediate termination of I/O, Oracle recommends that you precede the ALTER SYSTEM SUSPEND statement with a BEGIN BACKUP statement so that the tablespaces are placed in backup mode.

You must use conventional user-managed backup methods to back up split mirrors. RMAN cannot make database backups or copies because these operations require reading the datafile headers. After the database backup is finished or the mirrors are re-silvered, then you can resume normal database operations using the ALTER SYSTEM RESUME statement.

Backing up a suspended database without splitting mirrors can cause an extended database outage because the database is inaccessible during this time. If backups are taken by splitting mirrors, however, then the outage is nominal. The outage time depends on the size of cache to flush, the number of datafiles, and the time required to break the mirror.

Note the following restrictions for the SUSPEND/RESUME feature:

  • In a RAC configuration, you should not start a new instance while the original nodes are suspended.

  • No checkpoint is initiated by the ALTER SYSTEM SUSPEND or ALTER SYSTEM RESUME statements.

  • You cannot issue SHUTDOWN with IMMEDIATE, NORMAL, or TRANSACTIONAL options while the database is suspended.

  • Issuing SHUTDOWN ABORT on a database that was already suspended reactivates the database. This prevents media recovery or crash recovery from hanging.

To make a split mirror backup in SUSPEND mode:

  1. Place the database tablespaces in backup mode. For example, to place tablespace users in backup mode enter:

    ALTER TABLESPACE users BEGIN BACKUP;
    

    If you are backing up all of the tablespaces for your database, you can instead use:

    ALTER DATABASE BEGIN BACKUP;
    
  2. If your mirror system has problems with splitting a mirror while disk writes are occurring, then suspend the database. For example, issue the following:

    ALTER SYSTEM SUSPEND;
    
  3. Check to make sure that the database is suspended by querying V$INSTANCE. For example:

    SELECT DATABASE_STATUS FROM V$INSTANCE;
    
    DATABASE_STATUS 
    ----------------- 
    SUSPENDED 
    
  4. Split the mirrors at the operating system or hardware level.

  5. End the database suspension. For example, issue the following statement:

    ALTER SYSTEM RESUME;
    
  6. Check to make sure that the database is active by querying V$INSTANCE. For example, enter:

    SELECT DATABASE_STATUS FROM V$INSTANCE;
    
    DATABASE_STATUS 
    ----------------- 
    ACTIVE 
    
  7. Take the specified tablespaces out of backup mode. For example, enter the following to take tablespace users out of backup mode:

    ALTER TABLESPACE users END BACKUP;
    
  8. Copy the control file and archive the online redo logs as usual for a backup.

    Caution:

    Do not use the ALTER SYSTEM SUSPEND statement as a substitute for placing a tablespace in backup mode.

    See Also:

Making User-Managed Backups to Raw Devices

A raw device is a disk or partition that does not have a file system. In other words, a raw device can contain only a single file. Backing up files on raw devices poses operating system specific issues. The following sections discuss some of these issues on UNIX, Linux, and Windows.

Backing Up to Raw Devices on Linux and UNIX

When backing up to or from raw devices, the dd command on Linux and UNIX is the most common backup utility. See your operating system-specific documentation for complete details about this utility.

Using dd effectively requires specifying the correct options, based on your database. Table 27-1 lists details about your database that affect the options you use for dd.

Table 27-1 Aspects of the Database Important for dd Usage

Data Explanation

Block size

You can specify the size of the buffer that dd uses to copy data. For example, you can specify that dd should copy data in units of 8 KB or 64 KB. Note that the block size for dd need not correspond to either the Oracle block size or the operating system block size: it is merely the size of the buffer used by dd when making the copy.

Raw offset

On some systems, the beginning of the file on the raw device is reserved for use by the operating system. This storage space is called the raw offset. Oracle should not back up or restore these bytes.

Size of Oracle block 0

At the beginning of every Oracle file, the operating system-specific code places an Oracle block called block 0. The generic Oracle code does not recognize this block, but the block is included in the size of the file on the operating system. Typically, this block is the same size as the other Oracle blocks in the file.


The information in Table 27-1 enables you to set the dd options specified in Table 27-2.

Table 27-2 Options for dd Command

This option ... Specifies ...

if

The name of the input file, that is, the file that you are reading.

of

The name of the output file, that is, the file to which you are writing.

bs

The buffer size used by dd to copy data.

skip

The number of dd buffers to skip on the input raw device if a raw offset exists. For example, if you are backing up a file on a raw device with a 64 KB raw offset, and the dd buffer size is 8 KB, then you can specify skip=8 so that the copy starts at offset 64 KB.

seek

The number of dd buffers to skip on the output raw device if a raw offset exists. For example, if you are backing up a file onto a raw device with a 64 KB raw offset, and the dd buffer size is 8 KB, then you can specify skip=8 so that the copy starts at offset 64 KB.

count

The number of blocks on the input raw device for dd to copy. It is best to specify the exact number of blocks to copy when copying from raw device to file system, otherwise any extra space at the end of the raw volume that is not used by the Oracle datafile is copied to the file system.

Remember to include block 0 in the total size of the input file. For example, if the dd block size is 8 KB, and you are backing up a 30720 KB datafile, then you can set count=3841. This value for count actually backs up 30728 KB: the extra 8 KB are for Oracle block 0.


Because a raw device can be the input or output device for a backup, you have four possible scenarios for the backup. The possible options for dd depend on which scenario you choose, as illustrated in Table 27-3.

Table 27-3 Scenarios Involving dd Backups

Backing Up from ... Backing Up to ... Options Specified for dd Command

Raw device

Raw device

if, of, bs, skip, seek, count

Raw device

File system

if, of, bs, skip, count

File system

Raw device

if, of, bs, seek

File system

File system

if, of, bs


Backing Up with the dd utility on Linux and UNIX: Examples

For these examples of dd utility usage, assume the following:

  • You are backing up a 30720 KB datafile.

  • The beginning of the datafile has a block 0 of 8 KB.

  • The raw offset is 64 KB.

  • You set the dd block size to 8 KB when a raw device is involved in the copy.

In the following example, you back up from one raw device to another raw device:

% dd if=/dev/rsd1b of=/dev/rsd2b bs=8k skip=8 seek=8 count=3841

In the following example, you back up from a raw device to a file system:

% dd if=/dev/rsd1b of=/backup/df1.dbf bs=8k skip=8 count=3841

In the following example, you back up from a file system to a raw device:

% dd if=/backup/df1.dbf of=/dev/rsd2b bs=8k seek=8

In the following example, you back up from a file system to a file system, and so can set the block size to a high value to boost I/O performance:

% dd if=/oracle/dbs/df1.dbf of=/backup/df1.dbf bs=1024k

Backing Up to Raw Devices on Windows

Like Linux and UNIX, Windows supports raw disk partitions in which the database can store datafiles, online logs, and control files. Each raw partition is assigned either a drive letter or physical drive number and does not contain a file system. As in Linux and UNIX, each raw partition on Windows is mapped to a single file.

Windows differs from Linux and UNIX in the naming convention for Oracle files. On Windows, raw datafile names are formatted as follows:

\\.\drive_letter:
\\.\PHYSICALDRIVEdrive_number

For example, the following are possible raw filenames:

\\.\G:
\\.\PHYSICALDRIVE3

Note that you can also create aliases to raw filenames. The standard Oracle database installation provides a SETLINKS utility that can create aliases such as \\.\Datafile12 that point to filenames such as \\.\PHYSICALDRIVE3.

The procedure for making user-managed backups of raw datafiles is basically the same as for copying files on an Windows file system, except that you should use the Oracle OCOPY utility rather than the Windows-supplied copy.exe or ntbackup.exe utilities. OCOPY supports 64-bit file I/O, physical raw drives, and raw files. Note that OCOPY cannot back up directly to tape.

To display online documentation for OCOPY, enter OCOPY by itself at the Windows prompt. Sample output follows:

Usage of OCOPY:
     ocopy from_file [to_file [a | size_1 [size_n]]]
     ocopy -b from_file to_drive
     ocopy -r from_drive to_dir

Note the important OCOPY options described in the following table.

Table 27-4 OCOPY Options

Option Action

b

Splits the input file into multiple output files. This option is useful for backing up to devices that are smaller than the input file.

r

Combines multiple input files and writes to a single output file. This option is useful for restoring backups created with the -b option.


Backing Up with OCOPY: Example

In this example, assume the following:

  • Datafile 12 is mounted on the \\.\G: raw partition.

  • The C: drive mounts a file system.

  • The database is open.

To back up the datafile on the raw partition \\.\G: to a local file system, you can run the following command at the prompt after placing datafile 12 in backup mode:

OCOPY "\\.G:" C:\backup\datafile12.bak

Specifying the -b and -r Options for OCOPY: Example

In this example, assume the following:

  • \\.\G: is a raw partition containing datafile 7

  • The E: drive is a removable disk drive.

  • The database is open.

To back up the datafile onto drive E:, you can execute the following command at the Windows prompt after placing datafile 7 in backup mode:

# first argument is filename, second argument is drive
OCOPY -b "\\.\G:" E:\

When drive E: fills up, you can use another disk. In this way, you can divide the backup of datafile 7 into multiple files.

Similarly, to restore the backup, take the tablespace containing datafile 7 offline and run this command:

# first argument is drive, second argument is directory
OCOPY -r E:\ "\\.\G:"

Making Backups with the Volume Shadow Copy Service (VSS)

Volume Shadow Copy Service (VSS) is a set of Windows APIs that enable applications to create consistent snapshots called shadow copies. The Oracle VSS writer runs as a service on Windows systems and is integrated with VSS-enabled applications. You can use these applications to create snapshots of database files managed by the Oracle instance. For example, you can make shadow copies of an Oracle database while it is open read/write.

See Also:

Oracle Database Platform Guide for Microsoft Windows to learn how to back up and recover the database with VSS-enabled applications

Verifying User-Managed Datafile Backups

You should periodically verify your backups to ensure that they are usable for recovery.

Testing the Restore of Datafile Backups

The best way to test the usability of datafile backups is to restore them to a separate host and attempt to open the database, performing media recovery if necessary. This option requires that you have a separate host available for the restore procedure.

See Also:

"Performing Complete Database Recovery" to learn how to recover files with SQL*Plus

Running the DBVERIFY Utility

The DBVERIFY program is an external command-line utility that performs a physical data structure integrity check on an offline datafile. Use DBVERIFY primarily when you need to ensure that a user-managed backup of a datafile is valid before it is restored or as a diagnostic aid when you have encountered data corruption problems.

The name and location of DBVERIFY is dependent on your operating system. For example, to perform an integrity check on datafile users01.dbf on Linux or UNIX, run the dbv command as follows:

% dbv file=users01.dbf

Sample dbv output follows:

DBVERIFY - Verification starting : FILE = users01.dbf


DBVERIFY - Verification complete

Total Pages Examined         : 250
Total Pages Processed (Data) : 1
Total Pages Failing   (Data) : 0
Total Pages Processed (Index): 0
Total Pages Failing   (Index): 0
Total Pages Processed (Other): 2
Total Pages Processed (Seg)  : 0
Total Pages Failing   (Seg)  : 0
Total Pages Empty            : 247
Total Pages Marked Corrupt   : 0
Total Pages Influx           : 0

See Also:

Oracle Database Utilities to learn about DBVERIFY