COBOL Manual for TNS and TNS/R Programs

Fault-Tolerant Processes
HP COBOL Manual for TNS and TNS/R Programs—522555-006
32-4
Process Pairs
After you have done these things, the fault-tolerant facility does the remainder of the
work to assure fault-tolerant operation.
Topics:
•
Process Pairs
•
Checkpointing
•
Using the Fault-Tolerant Facility
•
Designing Programs for the Fault-Tolerant Facility
•
Debugging Programs That Use the Fault-Tolerant Facility
The information in these topics is general. For specific details on checkpointing, see
the Guardian Programmer’s Guide. For details on the STARTBACKUP and
CHECKPOINT statements, see STARTBACKUP and CHECKPOINT.
Process Pairs
A process is the basic executable unit known to the operating environment—the
execution (in a processor) of a program. Specifically, the term program indicates a
group of instruction codes and initialized data—an HP COBOL run unit; the term
process denotes the changing states of an executing program. The same loadfile can
be executing concurrently a number of times, but each execution is a separate
process.
An application process can be designed to recover from any type of hardware failure
except one—a failure of the processor in which it is executing. One way to provide fault
tolerance is to establish the process as a process pair. A process pair consists of two
executions of the same loadfile: the primary process executes in one processor; the
backup process executes in another. Control in the program indicates whether the
process is executing in the primary mode to perform its task or in the backup mode to
monitor the primary process.
Figure 32-1. Process Pair
A
A'
Checkpoint Messages
System Messages
Operating
System
Primary
Process
(performs the work)
Backup
Process
(monitors the primary process)
VST716.vsd