The CPF Recovery File is a BDAM disk file that saves transmitted commands until a response is received from a remote machine. Only commands selecting or defaulting to WAIT(NO) are saved for re-transmission. Commands targeted only for the local machine are not saved.
When a TSS command with WAIT(NO) is entered CPF saves a command image on the CPF Recovery File before transmitting it over the link. When a response from the remote is returned CPF deletes the command from the file. If the response is not received CPF scans the Recovery File at the resumption of service and select all commands not responded and retransmits them. When a response is received, the command(s) are deleted from the file.
CPF scans the CPF recovery file:
Before you can use the Recovery File, it must be formatted through TSSMAINT. A DD statement must be inserted into the CA Top Secret jobstream to define the CPF Recovery File. For example:
//CPFFILE DD DSN=SYS2.TSS50.CPF.RECOVERY,DISP=OLD
Important! CPF Recovery File cannot be shared across multiple systems.
If the CPF Recovery File is not defined, command routing through CPF can still occur but there is no re-transmission of unresponded commands.
If the CPF Recovery File is full a message is written to the job CA Top Secret LOG and console each time CPF wants to write a message to the file but cannot. The CPF operation continues but, in case of failure, the unwritten command cannot be recovered.
You can remove pending commands from the CPF Recovery File by date and by node. This is useful to selectively remove commands (for example, when testing systems that are only occasionally active).
Examples: remove pending CPF commands
This example removes all records for the specified node:
TSS REMOVE(*CPFRECV) CPFNODE(xxxxx,yyyyy)
This example removes all records up to, and including, the specified date:
TSS REMOVE(*CPFRECV) UNTIL(mm/dd/yy)
|
Copyright © 2014 CA Technologies.
All rights reserved.
|
|