Topics | How To - Oracle on Unix | How To - Oracle on Windows
Determine Required Archive Files for Data Recovery
Recover Oracle Data Using Media Explorer on Windows
Required Capability: none
To determine which archive files are needed to complete a data recovery:
If the data being recovered is of File System type and only part of the data is required or the entire Database application was secured by a data protection operation of the default subclient, then the number of subclients required can be as few as one.
If the entire File System is required or the data being recovered is a database application, then most likely all subclients that contained any part of this application will be required.
This identification process is greatly facilitated if you had run Media Information and CommCell Configuration reports while the CommServe was still available, and have those reports available as reference materials. |
(At which option your stop is dependent upon your particular recovery scenario.)
Before You Begin
Regularly running the above reports prior to the need for a disaster recovery will facilitate your identification and location of media archive files. |
To Catalog the Media using Media Explorer:
It is possible to perform a recovery using only
the default archive database. However, if recovering large numbers of
or many types of archive files makes locating files cumbersome, you
can create additional database file paths using the New button. The New and Open buttons can be used to start cataloging to a new catalog or to open an existing catalog database. This is useful in situations where you would like to use Media Explorer over multiple sessions and not overwrite previous catalog operations. |
Use the Media Information and CommCell Configuration reports to assist you in determining from which media it will be necessary to restore. If you do not have any reports saved, you will have to manually determine the required media. This involves finding the correct (latest) archive file(s) for backups you want to restore.
A full DLT 7000 tape will take approximately two hours to catalog. Be certain you have enough disc space on your local computer to accommodate the files. |
The CV_MAGNETIC
folder structure represents the top level of the magnetic mount path. The V_nn folder represents a volume which is equivalent to a storage policy copy. The AF_nnn folders represent the individual archive files within a volume. Each backup creates one or more archive files, depending on the iDataAgent. |
An archive file will sometimes span across several tapes during the course of normal backups. In this case when a catalog operation is run on the first media, the information on the second media may not appear to have been cataloged. However, when you perform a recovery from the first tape, Media Explorer asks you to load the second tape manually to continue the recovery. |
In some cases, a data protection operation will start on one media and complete on another. This we refer to as a spanned job . At most there can be one spanned job per any given media.
If after cataloging a media you see an archive file for which the Catalog column does not display Found All and the Catalog State column displays Partial, this indicates the archive file is spanned to another media. In this case, you need to locate that other media and catalog it as well. Once the last media that contains the spanned archive has been reached, you will be prompted to reinsert all the media (starting with the first) necessary to finish the cataloging process for the spanned job.
Once all media has been successfully cataloged, the Catalog column should display Found All and the Catalog State column should display Complete, and the recovery may take place like any other recovery. During the recovery of this archive file, you will be prompted to insert the media as required to complete the recovery.
Media Explorer stores cataloged media information in its database on your local hard drive for future reference. Once your files have been cataloged, you can go back and obtain this information without running the catalog operation again. To have the information available on another computer, however, you would need to run a catalog on that computer. |
You can click View Chunks from the File menu to display the Chunk Details window. This window provides Media Explorer database details that may be helpful to you.
Recovery scenarios include the following:
Consider the following:
Archive File TypesThe Oracle Database iDataAgent data protection operations create two types of archive files:
Oracle Media Explorer supports a single stream for tape libraries and multiple streams for magnetic libraries.
Data EncryptionWhen restoring data that was encrypted using a pass-phrase, you must provide the pass-phrase that was in use at the time the data protection operation was run. If you do not provide the pass-phrase, the restore will hang and a message will appear in the DrRecovery Log.
Before You Begin
To recover Oracle databases using Media Explorer on Windows
Make sure ORACLE_SID is set to the database that you are recovering. |
rman target <target connect string> catalog <catalog connect string>
a) To perform a full restore of a database:
run{
allocate channel ch1 type 'sbt_tape';
restore database;
recover database;
sql "alter database open";
}
b) To restore using multiple streams:
run{
allocate channel ch1 type 'sbt_tape';
allocate channel ch2 type `sbt_tape';
restore tablespace 'SYSTEM','USERS';
recover database;
sql "alter database open";
}
c) If data was encrypted using a pass-phrase, you must provide the pass-phrase that was in use at the time the data protection operation was run. Add the following parameter into the RMAN script as follows:
run{
allocate channel ch1 type `sbt_tape'
PARMS="ENV=(CvPswdPhrase=<yourpassphrase>)"TRACE 0;
restore database;
recover database;
}