prematurely or abnormally terminates. It does not
process to normal end of job (EOJ). When this occurs,
the operator is expected to take whatever corrective
actions are necessary to get the job going again. More
often than not, the operator is able to recover a job by
recreating a tape/disk file, moving the file to another
device, or possibly cleaning the read/write mechanisms
of the device prior to rerun. But, there are times when
the operator will notify you (the I/O control clerk) to
assist in correcting the problem. Such would be the case
when the input parameters are in error, the users input
is bad, or the job aborted because of an unrecoverable
program error. If this happens, you maybe responsible
for collecting all the data, both input and output, along
with any memory dumps, and forwarding them all to
the programmer.
During the recovery phase of an operation, the oper-
ator may need you to provide certain input parameters
or tape/disk files before the job can be executed.
Because of time constraints, a job that abnormally
terminates may have to be rescheduled. If so, you may
be responsible for seeing to it that the job gets
rescheduled and that the user is notified of any job delay.
We could go on and on, but by now you are beginning
to get the picture. These examples are just a few of the
many things that can get in the way of achieving a
normal EOJ. We bring them to your attention to make
you aware of the types of problems that can and do arise,
and the manner in which you are to respond. Hopefully,
you now know and are aware that monitoring a job
means more than just calling up the operator to see how
the job is progressing. It means you must oversee the
job to its completion, doing whatever is necessary to
help keep the job (or system) on track.
Output Products
Output from computer processingThe work that
has been completedmay take the form of a printed
document, magnetic tape, or magnetic disk or diskette.
In all cases, both you and the computer operator are
responsible for ensuring that all completed jobs run
successfully. In addition, you are responsible for
identifying and coordinating the various outputs for
each job, and for initiating their correct distribution.
To determine whether a job (or system) ran
successfully (to a normal EOJ) and that all processing
steps were properly performed, you may have to review
the computer console printout. This printout indicates
such things as the number of input records read, the
various input files updated, all error conditions (error
messages) that the operator encountered during the run
and the resulting actions taken, the various output files
created, and so on.
In the majority of cases, the computer console print-
out will provide you with the answers you are looking
for when it comes to reconciling processing discrepancies.
For example, it will inform you of the reasons certain
output productstapes, diskettes, or report listings
were not produced. Possibly the operator selected an
incorrect program option, or the input parameters were
incorrect or incomplete before starting the job. In short,
you are responsible and also accountable for every job
you work on, from the time it is submitted by the user
until its delivery back to the user.
When checking the users output, you should once
again refer to the run sheet and/or task folder to verify
that all items requested were, in fact, produced. If the
output is in the form of magnetic tape, disk, or diskette,
be sure it is labeled properly, given the proper
classification, and it is on the appropriate media
(magnetic media that has been designated for mail-out
or distribution only).
When checking reports, make sure they were run
on the proper forms (size and type), that no pages are
missing and the correct number of copies were printed,
and that all print is legible and lined up properly.
Once the output is checked, you then package each
completed copy of the report, along with any other out-
put products and the original input, place it in the proper
pickup area, and log the job out in the job control log.
You may need to notify the user when the job is ready.
If, during the course of checking over the users
output, you happen to come across something unusual
or you find an error, by all means, pull (reject) the job
immediately, bring it to the attention of your superior,
and notify the user of the delay. Even at this late stage,
it is better to reject a job to correct any problems or
discrepancies rather than to release it, only to have it
returned for rerun later.
USER SUPPORT
The term user support covers a broad range of duties.
They include answering inquiries from users, providing
logistical support, and processing trouble reports.
User Inquiries
Normal inquiries from users include system status,
job status, and reporting trouble. It is the job of the
technician to answer these questions promptly and
accurately. A user might ask:
l Why is the system slow?
1-7