HP NonStop RDF, Руководство по управлению

Страница: 200 / 478

NOTE:
You should only delete the TAKEOVER trigger on the backup system prior to issuing
the TAKEOVER command. If you delete the TAKEOVER trigger on the backup system when
you are not intending to execute a takeover operation, then you must remember to delete the
trigger on the primary system too when the latter comes back online. Failure to do this means
that when you start RDF next, RDFCOM will copy the TAKEOVER trigger information over to
the backup system, thereby reinstating it on that backup system.
Security Restrictions
You can issue the DELETE command if you are a member of the super-user group.
RDF State Requirement
After RDF is initialized, you can issue a DELETE command only when RDF is stopped.
Usage Guidelines
For the DELETE command to have any effect, a configuration record must already exist for the
secondary image trail or updater process associated with the volume name supplied (that is,
someone must have previously issued an ADD IMAGETRAIL or ADD VOLUME command for
the volume).
When you issue a DELETE VOLUME command, RDF responds:
• The extractor process stops sending image data for the volume specified in the DELETE
VOLUME command.
• The updater process associated with this volume will not be started.
Use the DELETE VOLUME command if an update volume on the backup system becomes
unusable and you want RDF to continue maintaining the other volumes. In such a case, you
must stop RDF at the primary system, issue the DELETE VOLUME command, and then restart
RDF.
When it is convenient to do so, you can resynchronize the affected volume, configure a new
updater process by issuing appropriate SET VOLUME commands, and then issue an ADD
VOLUME command to restart RDF protection for the affected primary volume.
Before you can remove an image trail with the DELETE IMAGETRAIL command, you must
delete all updater processes that are configured to that image trail. The DELETE IMAGETRAIL
command then deletes the configuration record for the image trail and all image files currently
belonging to that trail. Thus, the network connection to the backup system must be available
when you enter this command.
If you need to change an updater’s image trail volume, it is recommended that you stop TMF;
wait for RDF to stop; delete the volume; re-add the volume back into the configuration, associating
it to a different image trail volume; and then restart the TMF and RDF subsystems. This is the
only way to ensure that the backup database will remain synchronized with the primary database.
Examples
Assume that RDF is protecting primary system data volumes $DATA01, $DATA02, and $DATA03,
and that all three volumes are configured to the MAT. Assume also that the changes are being
replicated to backup system volumes $DATA1, $DATA2, and $DATA3, and that the updaters
for those volumes are acquiring their audit data from secondary image trail volumes $SECIT1,
$SECIT2, and $SECIT3, respectively.
To delete the configuration records for the updater process and secondary image trail associated
with $DATA03, enter the following commands:
]DELETE VOLUME $DATA03
]DELETE IMAGETRAIL $SECIT3
200 Entering RDFCOM Commands