This document contains special considerations that refer to Natural in batch mode under the operating system z/OS.
The section covers the following topics:
For considerations that refer to Natural in batch mode generally, see also:
The Natural z/OS batch interface consists of the NATOS
object module which is linked to the Natural nucleus during the
installation
procedure for base Natural as described in the
Installation for z/OS documentation.
You can customize the Natural z/OS batch interface to meet your
requirements by changing the parameter settings in the
NTOSP
macro in the Natural parameter module during the
appropriate
installation step.
NATOS
is fully reentrant and can run above the 16 MB
line. Multiple Natural sessions can be started in parallel within one batch
region; see
Subtasking
Session Support for Batch Environments.
For information on the driver parameters that are available for z/OS
in batch mode, refer to the description of profile parameter
OSP
or
parameter macro NTOSP
in the Parameter Reference documentation.
The following data sets are required if certain functions are used during a Natural z/OS batch mode session:
Data Set | Explanation |
---|---|
CMEDIT
|
Software AG Editor Work File |
CMHCOPY
|
Hardcopy Print Output |
CMOBJIN
|
Input for Natural INPUT Statements
|
CMPLOG
|
Dynamic Profile Parameter Report Output |
CMPRINT
|
Primary Report Output |
CMPRMIN
|
Dynamic Profile Parameter Input |
CMPRTnn
|
Additional Reports 01-31 |
CMSYNIN
|
Primary Command Input |
CMTRACE
|
External Trace Output |
NATRJE
|
Job Submit Output |
STEPLIB
|
Load Library for External Modules |
CMWKFnn
|
Work Files 01-32 |
These data sets are described below.
For sequential data output sets, the default DCB
RECFM/LRECL
information is as follows:
RECFM=FBA
and LRECL=133
The Software AG Editor work file VSAM data set is required if a local or global Software AG editor buffer pool is to be used.
If not defined in the JCL, the name of the Editor work file
specified by subparameter DSNAME
of
profile parameter EDBP
or parameter
macro NTEDBP
is used by Natural to do the dynamic allocation for the Editor work file.
Alternatively, profile parameter
EDPSIZE
can be used to run with an auxiliary editor buffer pool, which doesn't require
an editor work file. For more information about the installation of the
Software AG editor, refer to
Installing the
Software AG Editor on z/OS in the Installation
for z/OS documentation.
The default name of the hardcopy print output data set is
CMHCOPY
. It can be changed by one of the following:
the subparameter DEST
of
profile parameter PRINT
for Print File
0,
the profile parameter HCDEST
, which is an
equivalent of PRINT=((0),DEST=...),
the setting of the system variable
*HARDCOPY
during the session,
the terminal command %H
during the
session.
The subparameters of the profile parameter
PRINT
for
Print File 0 can be used to change the default values for the hardcopy data
set. The default data set name CMHCOPY
implies
CLOSE
=FIN
for the hardcopy print data set, that is, after the data set has been opened
for output, any subsequent change of the hardcopy print output data set name
will not be honored. If a different name is defined at open time, the hardcopy
data set will be closed according to subparameter
CLOSE
of
profile parameter PRINT
for Print File
0.
During the session, the hardcopy data set can be released and
reallocated (before open or after close) by the by dynamic allocation (via
application programming interface USR2021N
, see
SYSEXT - Natural Application
Programming Interfaces).
This data set can be used to read data by the Natural
INPUT
statement rather
than from the primary input data set CMSYNIN
.
The usage of CMOBJIN
is controlled by the profile
parameter OBJIN
. The input
record data length for Natural is determined by profile parameter
SL
. The
maximum record length (LRECL
) supported is 255. The record format
(RECFM
) can be fixed or variable.
If profile parameter PLOG
=ON
is set and
data set CMPLOG
is available, the evaluated dynamic profile
parameters are written to this data set during session initialization. If data
set CMPLOG
is not available, the evaluated dynamic profile
parameters are written to CMPRINT
.
CMPRINT
is used for the primary output report
resulting from DISPLAY
,
PRINT
and
WRITE
statements in a
Natural program.
The record format (RECFM
) for CMPRINT
is
FBA. If no DCB information for LRECL
is available from the data
set or from the JCL, LRECL=133
will be used as default. If no DCB
information for BLKSIZE
is available from the data set or from the
JCL, BLKSIZE
will be 10 times the value of LRECL
as
default.
If not defined in JCL, CMPRINT
will be allocated
dynamically as
//CMPRINT DD SYSOUT=*
when the first record is to be written.
CMPRMIN
can be used as a
dynamic
parameter data set to overcome the length restriction for the
character string in the job control PARM
keyword of the
EXEC
statement.
If available, this file is read during session initialization to get the dynamic profile parameters.
All input records from CMPRMIN
are concatenated into
one parameter string. Only the first 72 positions of each CMPRMIN
record are significant. Trailing blanks at the end of each record are
truncated; if the last non-blank character is a comma, all trailing blanks are
truncated, otherwise just one blank is left as delimiter; no commas are
inserted.
Additional dynamic parameters can be supplied using the job
control PARM
keyword. They are concatenated at the end
of the parameter string which was built from the input of CMPRMIN
,
that is, these can be used to overwrite the parameters from
CMPRMIN
.
A comment starts with a slash asterisk /*
and ends
with an asterisk slash */
and can be anywhere in your data set.
They may span over several lines in your file.
These data sets can be used by Natural print file statements like
WRITE
(nn
). If no DCB information (for
example, RECFM
, LRECL
, BLKSIZE
) is
available, the defaults are defined by the PRINT
profile
parameter or the NTPRINT
macro in the Natural parameter module. The print file names can be overwritten
by subparameter DEST
.
This data set is used to read command input and data requested by
the Natural INPUT
statement. The latter is controlled by the profile parameter
OBJIN
(see
also CMOBJIN
).
The input record data length for Natural is determined by profile
parameter SL
.
The maximum record length (LRECL
) supported is 255. The record
format (RECFM
) can be fixed or variable.
If profile parameter ETRACE
=ON
is set
or the equivalent terminal command %TRE+
was issued,
any Natural trace output during the session is written to the
CMTRACE
data set. To define the Natural components that are to be
traced, the profile parameter TRACE
is
required.
If data set CMTRACE
is not available, it will be
allocated dynamically as
//CMTRACE DD SYSOUT=*
when the first trace record is to be written.
This data set is used for the Natural job submitting utility. If it is not defined, it will be allocated dynamically as
//NATRJE DD SYSOUT=(A,INTRDR)
when the first job is submitted.
STEPLIB
is the default load library name for loading
external modules, for example:
the environment-independent nucleus (profile parameter
NUCNAME
),
a separate Adabas link routine module (profile parameter
ADANAME
),
the session back-end program (profile parameter
PROGRAM
),
any external subprograms not linked to the Natural parameter module.
The load library name can be changed by profile parameter
LIBNAM
.
The specified load library name must be defined by a DD statement in the
JCL.
These data sets can be used by Natural work file statements like
READ WORK
nn
and
WRITE WORK
nn
.
If no DCB information (RECFM
, LRECL
,
BLKSIZE
, etc.) is available in the JCL or in the VTOC entry for
the data set, the defaults are defined by the WORK
profile
parameter or the NTWORK
macro in the Natural parameter
module.
The work file data set names can be overwritten by subparameter
DEST
.