Mirroring Databases
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Creating Volumes
Remote Mirroring of an Oracle Database Using Data Replicator Software
3-9
·
Use RollForward to recover the database at the secondary site.
·
Suspend secondary-site RollForward recovery to create additional recovery points.
·
Control the primary-site RDBMS environment to increase transaction log mirroring
during contingency situations.
·
Make sure that the primary site has persisted data and has mirrored the data. Make sure
that the primary site is shut down before switching to the secondary site that has the
fully-recovered copy of the database.
Mirroring Databases
Mirroring an entire database is not complicated. The process begins with identifying those
RDBMS objects that are required for a complete restart.
The following RDBMS objects are required:
·
Tablespaces for the complete database
·
Indexes
·
Transaction logs
·
Control files for RDBMS operation
·
Password files for user logon IDs
·
RDBMS home directory
·
Initialization files
·
Supporting files, such as utilities or report writers
·
Troubleshooting tools
The following RDBMS objects are not required to be mirrored:
·
Swap files for the RDBMS and the operating system (OS)
·
Sort directories and associated disk storage
The required RDBMS objects that are candidates for mirroring are those that are
absolutely required for a complete recovery at the secondary site.
Creating Volumes
Either through SANshare Storage Partitioning or the command line interface utility, create
the volumes on the remote storage system that will contain the mirror data from the
primary site. These volumes should be similarly named as the primary volumes to allow
for easy identification during the next step.
