RDF System Management Manual for J-series and H-series RVUs (RDF 1.10)

• RDFVOLUME
• EXTENTS
• FASTUPDATEMODE
The ATINDEX attribute specifies an integer value identifying a configured TMF audit trail on the
primary system. 0 specifies the MAT. 1 through 15 specify auxiliary audit trails AUX01 through
AUX15. The default is 0. For each configured extractor, there must be a corresponding receiver
with the same ATINDEX value. For information about protecting auxiliary audit trails, see Chapter 13
(page 276).
The CPUS attribute specifies the processors in the backup system in which the receiver is to run.
The PRIORITY attribute specifies the priority at which the receiver will run. You should set the
receiver’s priority higher than that of any application’s process and higher than that of any RDF
updater process.
The PROCESS attribute supplies a name for the receiver process. You should specify a meaningful
mnemonic such as $RECV. The process name can be any unique valid process name up to six
characters, including the $ symbol. However, you cannot specify HP reserved process names that
are of the form $X*, $Y*, or $Z*, in which * is any alphanumeric string.
The RDFVOLUME attribute applies only to the master receiver. It specifies which volume on the
backup system will contain the receiver’s master image trail. The file naming convention for image
trail files is $volume.control-subvolume.AAnnnnnn, where n is a digit. For example, the
first image file is named $volume.control-subvolume.AA000001. You cannot specify the
subvolume name because that name is controlled by RDF.
The EXTENTS attribute only applies to the master receiver. It specifies the size of the primary and
secondary extents for all image trail files on all image trails.
The FASTUPDATEMODE value controls the frequency with which the receiver writes to the image
trails and makes image trail data available to the updaters. With FASTUPDATEMODE OFF, the
receiver buffers the audit sent by the extractor and writes those buffers out to the image trails at
the most convenient time. This ensures that RDF can achieve the highest possible extractor-to-receiver
throughput, but it does delay the updaters in how quickly they are allowed to read and apply the
audit to the backup database. One can typically observe updater RTD times in the range of 1-20
seconds, although it may only take an updater a fraction of one second to apply 20 seconds worth
audit.
With FASTUPDATEMODE ON, as a receiver receives an extractor message, it buffers all the audit
sent in that message by the extractor, writes those buffers immediately to the image trails, and then
makes that data immediately available to the updaters. Depending on the value of the
UPDATERDELAY attribute in the global RDF configuration record, the updaters can then read the
image trails and apply the freshly written audit to the backup database immediately, thereby
keeping updater RTD times to the lowest possible value. Because the receiver writes the audit
immediately to the image trails after processing each extractor message, having FASTUPDATEMODE
set ON can impact extractor-to-receiver throughput.
For a complete discussion of FASTUPDATEMODE, see the description involving the SET RECEIVER
command in Chapter 8 (page 176).
To configure an RDF receiver process named $RECV to run as a process pair in CPUs 0 and 2 of
the backup system at a priority of 185 with FASTUPDATEMODE off, and to have the RDF image
trail file (with a primary extent size of 3000 pages and a secondary extent size of 3000 pages)
reside on the volume $IMAGE, issue the following commands:
]SET RECEIVER ATINDEX 0
]SET RECEIVER PROCESS $RECV
]SET RECEIVER CPUS 0:2
]SET RECEIVER PRIORITY 185
]SET RECEIVER RDFVOLUME $IMAGE
86 Installing and Configuring RDF