HP NonStop RDF, Руководство по управлению

Страница: 342 / 478

can have up to 16 active, 16 restore, and 16 overflow volumes, an extractor list can contain up to
48 volume names. Set each volume name:
SET EXTRACTOR VOLUME volume-name
volume-name must be a valid volume name specified in the current TMF configuration on your
primary system. Use a SET statement for each individual volume. You do not need to specify
whether the volume is an active volume, restore volume, or overflow volume; you merely specify
the volume name. The list of volumes you specify is placed in the extractor configuration record.
When you ADD the extractor attributes, RDFCOM checks to ensure that each configured volume
is a valid volume. If any are not, the ADD command fails with an error for the first volume name
that is either an invalid name or does not correspond to a valid volume.
ALTER RDF Remote Mirror Configuration
You can alter the RDF REMOTE MIRROR configuration attribute to turn ZLT protection on and
off as needed. If entered on the primary system, you must stop RDF, alter the attribute, and then
restart RDF. If the primary system is no longer available and you are preparing to start the RDF
takeover operation, you can alter the attribute on the backup system. The syntax is:
ALTER RDF REMOTE MIRROR [ ON | OFF ]
If you issue the ALTER command on the backup system while the network is down and the
primary system is still up, you must then issue the ALTER command again on the primary system
when the network comes back up; otherwise, the change will not be kept on either the primary
or the backup system.
If one of your remote mirrors fails during normal operations on your primary system, you might
want to turn the RDF REMOTE MIRROR configuration attribute off on the primary system so
that if you need to execute a takeover operation you can complete the takeover as quickly as
possible. Remember, if you have the RDF REMOTE MIRROR configuration attribute turned off
at the time of a takeover, you do not have ZLT protection. When the troublesome mirror comes
back up and is fully revived so that it is current with the local mirror, you then re-enable ZLT
on the primary to reinstate ZLT protection.
ZLT Takeover Operations
CAUTION:
At the time of a ZLT takeover, HP strongly recommends that before connecting the
remote mirrors to the standby system, you either disconnect the remote mirrors from the primary
RDF system, or disable the XP disk array Logical Unit (LUN) from the primary system. If the
remote mirror's LUN remains shared and active with both the primary and the standby RDF
systems, it might cause data corruption on the remote mirror disk.
If ZLT is configured and enabled when a TAKEOVER command is issued on the backup system,
the overall takeover operation executes in two phases.
NOTE:
Before issuing the TAKEOVER command, you must have connected the remote mirrors
to the standby system. When the remote mirrors are connected to the standby system, the audit
records on the remote mirrors have no relationship to the audit trail on the standby system. The
remote mirrors are not part of the TMF configuration of the standby system.
Phase 1 (ZLT Processing)
RDFCOM stops all RDF processes on the backup system. If the standby and backup systems are
not the same system, RDFCOM copies the RDF configuration file on the backup system to the
standby system. RDFCOM then starts an RDF monitor process on the backup system. That
monitor then starts the extractor(s) on the standby system and the receiver(s), purger, and
updater(s) on the backup system.
342 Zero Lost Transactions (ZLT)