SQL/MX 3.2.1 Management Manual (H06.26+, J06.15+)

others, are described next. For more information about volume recovery and file recovery, see the
TMF Operations and Recovery Guide.
Using TMF to Recover Dropped SQL/MX Objects
TMF is your primary means of recovering dropped SQL/MX objects. For specific recovery
procedures, see:
• “Recovering Range Partitions” (page 239)
• “Recovering SPJs” (page 240)
• “Recovering Tables” (page 240)
• “Recovering Views and Indexes” (page 251)
Using RDF to Set Up a Backup System
The RDF subsystem monitors changes made to a production database on a local (primary) system
and maintains a copy of that database on a remote (backup) system. Because it applies changes
to the backup database as soon as they are detected on the primary system, the RDF subsystem
keeps the backup database continuously up to date with changes made by business applications
on the primary system. You are able, therefore, to switch your business operations from the primary
system to the backup system with minimal interruption and loss of data in the event of unexpected
or planned outages of the primary system.
NOTE: If you run SQL DDL operations with Shared Access on RDF-protected objects, you must
be certain that the operation is committed on your primary system before you attempt to perform
the operation on your RDF backup system.
The RDF subsystem also enables you to use the backup database as a read-only resource to balance
the overall workload and improve response times. Activities at the backup system can include
querying the database, processing heavy batch-reporting loads, and consolidating data from
multiple sites into one central site. For more information, see the RDF/IMP, IMPX, and ZLT System
Management Manual.
RDF configuration enables users to map volumes on the primary node to different volumes on the
RDF backup node. Except for this volume-to-volume mapping, the rest of the primary and backup
file names must match exactly. This subsection describes how to create an RDF backup database
with the correct file names, populate the backup database, and synchronize it with the primary
database.
You can create RDF backup databases online or offline. Offline creation uses Backup and Restore
2 and requires that the primary database be taken offline until the backup database is created
and RDF replication has started. Online creation of the backup database allows the primary
database to remain online while the backup database is being created and populated.
Online creation uses several methods to create tables on the RDF backup node. These tables can
be populated using INSERT...SELECT. To use this technique, you must configure either the primary
or backup catalog as a remote catalog by using the SQL/MX distribution REGISTER CATALOG
command. Operations such as the INSERT...SELECT statement need to involve tables from both
catalogs. For more information, see the SQL/MX Reference Manual.
56 Planning Database Security and Recovery