Table of Contents
The linking of systems into a multihost network provides the capability for a type of distributed processing. Each process executes on a single host system. However, the various members of a process family can run on different host systems and can communicate with each other in most of the same ways they could if they were all running on the same system.
This type of distributed processing is referred to as remote tasking and is provided by Host Services software. Remote tasking is supported across BNA Version 2 and Open Systems Interconnection (OSI) networks
This section uses some specialized terminology to discuss remote tasking. The term remote process is used to refer to a process that is initiated from one host system, but runs on another host system. The host from which the remote process is initiated is referred to as the local host. The host at which the remote process runs is referred to as the remote host.
In the same way, the local operator is an operator at the system from which the remote process is initiated. The remote operator is an operator at the system where the remote process runs.
A remote process can be initiated from programs or from interactive sources such as the operator display terminal (ODT), a Command and Edit (CANDE) session, or a Menu-Assisted Resource Control (MARC) session. Any messages generated by the process are routed back to the local ODT and originating terminal. You can monitor and control the remote process by transmitting ODT commands to the remote host system.
The following are reasons why you might want to initiate a remote process:
-
To equalize the processor load on the various systems at an installation. If the local system is overloaded, a process may run more quickly at a remote host.
-
To make use of a program stored at a remote host. A process must run on the same system where the object code file is stored. Therefore, initiating a program that is stored at another system implies the creation of a remote process.
-
To efficiently access files stored on a remote host. A remote process running on the remote host can access these files more efficiently than a local process that accesses the files using Host Services logical I/O. The result can be savings in I/O time and elapsed time.
For further information about Host Services, other than remote tasking, refer to the Distributed Systems Services (DSS) Operations Guide. For additional information about Host Services logical I/O, refer to the I/O Subsystem Programming Guide.
