Job Entry Subsystem

from Wikipedia, the free encyclopedia

The Job Entry Subsystem ( JES ) is a subsystem under z / OS on IBM - mainframes .

In the early days of the IBM mainframe was the so-called stacking and batching the only possible form of data processing. Back then, so -called jobs were encoded on punched cards, read into the system via punched card readers and processed sequentially. Even if punch cards are no longer processed today, batch processing continues to play an important role. Today it is mainly used for automated processing, often for so-called mass processing. Back then, as now, the control instructions for batch processing are coded in the Job Control Language (JCL).

As a subsystem of the z / OS operating system, the Job Entry Subsystem controls the jobs to be executed. It is required by the operating system to receive, process, and schedule jobs, to forward them to the operating system for execution and to receive the job's output. In short, the JES brings the jobs into the operating system and receives the output of the jobs.

There are two job entry subsystems for z / OS: JES2 and JES3 , with JES2 being the most common.

Working method

The processing of a job by the JES is divided into three parts:

  1. Preprocessing:
    In this part the actual file is read and interpreted. If a syntax error occurs here, processing is terminated here. Files are also allocated; this may require support from the operator or from hierarchical storage management (HSM).
  2. Processing:
    This is where the actual steps are carried out and the programs started.
  3. Postprocessing:
    This is the "cleanup phase" of the job run: files that are no longer required are deleted, files that were blocked for exclusive processing are released again, and the output is printed.

variants

JES is available in two versions: JES2 and JES3.

In principle, both do the same thing, the differences for the user are minor:

  • JES3 offers the possibility to distribute jobs on different computers (nodes) or to create job networks .
  • JES2 is easier to handle in terms of configuration, which is why operators usually use JES2 on individual computers.

JES2

JES2 is the Job Entry Subsystem in version 2.

JES2 emerged from HASP (Houston Automatic Spool Program), which was developed by IBM for NASA in the 1960s . Even today, JES2 writes its messages as $ HASP messages in the system log.

JES2 is easier to set up than JES3, but can not use the advantages of clusters .

However, it is possible to instruct JES2 to run a job on a specific system. This happens with the following record:

/*JOBPARM SYSAFF=system

JES3

JES3 is the Job Entry Subsystem in version 3.

JES3 offers more options for controlling jobs, but the configuration effort is considerably higher than JES2. The defining feature are job networks .

A record, independent of // * NET, which is described in more detail under the job networks, is mentioned here:

//*MAIN ORG=dst,CLASS=class,SYSTEM=system

The individual parameters here are:

  • ORG = dst : The destination , which corresponds to the output address . Here you can enter certain values ​​predefined by the data center on which the output should take place. ORG = LOCAL, for example, could be the printer in the data center.
  • CLASS = class : A job class can be specified here, which helps the operators to classify jobs. This class hardly differs from the JCL parameter CLASS ; however, it is possible to specify several characters instead of just one.
  • SYSTEM = system : Here you can select the system on which the job is to run. The use of this parameter is not advisable (except in really justified exceptional cases), since the ability of JES3 to distribute jobs to different systems is avoided.