RDF/IMP and IMPX System Management Manual (RDF 1.3+)
Managing RDF
Compaq NonStop™ RDF/IMP and IMPX System Management Manual—522204-001
5-18
Issuing the TAKEOVER Command
•
If RDF had been stopped, RDFCOM starts the monitor in TAKEOVER mode. The
monitor starts a receiver process and all updater processes. (The name of this
monitor is generated by the system.)
In either state, each updater processes to the end of its RDF image trail. If an updater
reads an audit record but the outcome of the transaction is unknown (the corresponding
commit or abort record did not make it to the backup system), the updater does not apply
that record. Instead, it logs information about the record in that updater’s exception file
(each updater process has its own exception file).
Upon reaching the end of the last RDF image file, each updater process stops. When all
of the updater processes have stopped, the purger logs either a message 724 or 725 and
then stops. Message 724 indicates that the takeover completed successfully. Message
725 indicates that it did not, and you should reissue the TAKEOVER command. A
message 724 is always followed by a message 735, which indicates the last MAT
position seen by the receiver process.
If a monitor process was started on the backup system, the monitor stops after the other
RDF processes are shut down.
The following example shows how to use the TAKEOVER command when RDF is
running with updating enabled. The command sequence in the example causes the
backup system \TORONTO to take over processing from the primary system
\SANFRAN.
1. At the TACL prompt on the backup system (\TORONTO), start an RDFCOM
session and specify the control subvolume name:
>RDFCOM SANFRAN
2. At the RDFCOM prompt, issue the TAKEOVER command:
]TAKEOVER
RDF displays the following prompt message:
*** TAKEOVER assumes a disaster on \SANFRAN has occurred.
Are you sure you want to TAKEOVER?