|Oracle® Database Backup and Recovery User's Guide
11g Release 2 (11.2)
Part Number E10642-04
|PDF · Mobi · ePub|
This chapter explains how to restore and recover individual data blocks within a datafile. This chapter contains the following topics:
This section explains the purpose and basic concepts of block media recovery.
Lowers the mean time to recover (MTTR) because only blocks needing recovery are restored and recovered
Enables affected datafiles to remain online during recovery
Without block media recovery, if even a single block is corrupt, then you must take the datafile offline and restore a backup of the datafile. You must apply all redo generated for the datafile after the backup was created. The entire file is unavailable until media recovery completes. With block media recovery, only the blocks actually being recovered are unavailable during the recovery.
Block media recovery is most useful for physical corruption problems that involve a small, well-known number of blocks. Block-level data loss usually results from intermittent, random I/O errors that do not cause widespread data loss, as well as memory corruptions that are written to disk. Block media recovery is not intended for cases where the extent of data loss or corruption is unknown and the entire datafile requires recovery. In such cases, datafile media recovery is the best solution.
In most cases, the database marks a block as media corrupt and then writes it to disk when the corruption is first encountered. No subsequent read of the block will be successful until the block is recovered. You can perform block recovery only on blocks that are marked corrupt or that fail a corruption check.
If the database on which the corruption occurs is associated with a real-time query physical standby database, then the database automatically attempts to perform block media recovery. The primary database searches for good copies of blocks on the standby database and, if they are found, repairs the blocks with no impact to the query that encountered the corrupt block. The Oracle Database physical block corruption message (
ORA-1578) is displayed only if the database is unable to repair the corruption.
Whenever block corruption has been automatically detected, you can perform block media recovery manually with the
RECOVER ... BLOCK command. By default, RMAN first searches for good blocks in the real-time query physical standby database, then flashback logs and then blocks in full or level 0 incremental backups.
Note:For block media recovery to work automatically, the physical standby database must be in real-time query mode. An Oracle Active Data Guard license is required.
If a corrupt data block is discovered on a real-time query physical standby database, the server attempts to repair the corruption by obtaining a copy of the block from the primary database. The repair is performed in the background, enabling subsequent queries to succeed if the repair is successful. Automatic block repair is attempted if the following database initialization parameters are configured on the standby database as described:
LOG_ARCHIVE_CONFIG parameter is configured with a
DG_CONFIG list and a
LOG_ARCHIVE_DEST_n parameter is configured for the primary database with the
FAL_SERVER parameter is configured and its value contains an Oracle Net service name for the primary database
V$DATABASE_BLOCK_CORRUPTION view displays blocks marked corrupt by database components such as RMAN,
dbv, and SQL queries. The following types of corruption result in the addition of rows to this view:
Physical corruption (sometimes called media corruption)
The database does not recognize the block: the checksum is invalid, the block contains all zeros, or the block header is corrupt.
Physical corruption checking is enabled by default. You can turn off checksum checking by specifying the
NOCHECKSUM option of the
BACKUP command, but other physical consistency checks, such as checks of the block headers and footers, cannot be disabled.
The block has a valid checksum, the header and footer match, and so on, but the contents are logically inconsistent. Block media recovery may not be able to repair all logical block corruptions. In these cases, alternate recovery methods, such as tablespace point-in-time recovery, or dropping and re-creating the affected objects, may repair the corruption.
Logical corruption checking is disabled by default. You can turn it on by specifying the
LOGICAL option of the
The database can detect some corruptions by validating relationships between blocks and segments, but cannot detect them by a check of an individual block. The
V$DATABASE_BLOCK_CORRUPTION view does not record at this level of granularity.
Like datafile media recovery, block media recovery cannot generally survive a missing or inaccessible archived log, although it will attempt restore failover when looking for usable copies of archived redo log files, as described in "Restore Failover". Also, block media recovery cannot survive physical redo corruptions that result in checksum failure. However, block media recovery can survive gaps in the redo stream if the missing or corrupt redo records do not affect the blocks being recovered. Whereas datafile recovery requires an unbroken series of redo changes from the beginning of recovery to the end, block media recovery only requires an unbroken set of redo changes for the blocks being recovered.
Note:Each block is recovered independently during block media recovery, so recovery may be successful for a subset of blocks.
When RMAN first detects missing or corrupt redo records during block media recovery, it does not immediately signal an error because the block undergoing recovery may create a new one later in the redo stream. When a block is recreated, all previous redo for that block becomes irrelevant because the redo applies to an old incarnation of the block. For example, the database creates a new a block when users drop or truncate a table and then use the block for other data.
Assume that media recovery is performed on block 13 as depicted in Figure 19-1.
Figure 19-1 Performing RMAN Media Recovery
After block recovery begins, RMAN discovers that change 120 is missing from the redo stream, either because the log block is corrupt or because the log cannot be found. RMAN continues recovery assuming that block 13 will be recreated later in the redo stream. Assume that in change 140 a user drops the table
employees stored in block 13, allocates a new table in this block, and inserts data into the new table. At this point, the database formats block 13 as a new block. Recovery can now proceed with this block even though some redo preceding the recreation operation was missing.
The following prerequisites apply to the
RECOVER ... BLOCK command:
The target database must run in
ARCHIVELOG mode and be open or mounted with a current control file.
If the target database is a standby database, then it must be in a consistent state, recovery cannot be in session, and the backup must be older than the corrupted file.
The backups of the datafiles containing the corrupt blocks must be full or level 0 backups and not proxy copies.
If only proxy copy backups exist, then you can restore them to a nondefault location on disk, in which case RMAN considers them datafile copies and searches them for blocks during block media recovery.
RMAN can use only archived redo logs for the recovery.
RMAN cannot use level 1 incremental backups. Block media recovery cannot survive a missing or inaccessible archived redo log, although it can sometimes survive missing redo records.
Flashback Database must be enabled on the target database for RMAN to search the flashback logs for good copies of corrupt blocks.
If flashback logging is enabled and contains older, uncorrupted versions of the corrupt blocks, then RMAN can use these blocks, possibly speeding up the recovery.
The target database must be associated with a real-time query physical standby database for RMAN to search the database for good copies of corrupt blocks.
Typically, block corruption is reported in the following locations:
Results of the
BACKUP ... VALIDATE command
Error messages in standard output
The alert log
User trace files
Results of the SQL commands
Results of the DBVERIFY utility
Third-party media management output
For example, you may discover the following messages in a user trace file:
ORA-01578: ORACLE data block corrupted (file # 7, block # 3) ORA-01110: data file 7: '/oracle/oradata/trgt/tools01.dbf' ORA-01578: ORACLE data block corrupted (file # 2, block # 235) ORA-01110: data file 2: '/oracle/oradata/trgt/undotbs01.dbf'
In the following procedure, you identify the blocks that require recovery and then use any available backup to restore and recover these blocks.
To recover specific data blocks:
Obtain the datafile numbers and block numbers of the corrupted blocks.
SELECT NAME, VALUE FROM V$DIAG_INFO;
Start RMAN and connect to the target database, which must be mounted or open.
SHOW ALL command to confirm that the appropriate channels are preconfigured.
RECOVER ... BLOCK command at the RMAN prompt, specifying the file and block numbers for the corrupted blocks.
The following example recovers two blocks.
RECOVER DATAFILE 8 BLOCK 13 DATAFILE 2 BLOCK 19;
You can also specify various options to control RMAN behavior. The following example indicates that only backups with the tag
mondayam will be used when searching for blocks. You could use the
FROM BACKUPSET option to restrict the type of backup that RMAN searches, or the
EXCLUDE FLASHBACK LOG option to restrict RMAN from searching the flashback logs.
RECOVER DATAFILE 8 BLOCK 13 DATAFILE 2 BLOCK 199 FROM TAG mondayam;
To recover all blocks logged in V$DATABASE_BLOCK_CORRUPTION:
Start SQL*Plus and connect to the target database.
V$DATABASE_BLOCK_CORRUPTION to determine whether corrupt blocks exist. For example, execute the following statement:
SQL> SELECT * FROM V$DATABASE_BLOCK_CORRUPTION;
Start RMAN and connect to the target database.
Recover all blocks marked corrupt in
The following command repairs all physically corrupted blocks recorded in the view:
RMAN> RECOVER CORRUPTION LIST;
After the blocks are recovered, the database removes them from
See Also:Oracle Database Backup and Recovery Reference to learn about the
RECOVER ... BLOCKcommand