The JOBVAR statement has two different formats, one for saving job variables and one for restoring job variables. The JOBVAR statement for restoring job variables is described here; the JOBVAR statement for saving job variables is described in the section "JOBVAR Select job variables in save runs".
In the JOBVAR statement the job variables to be read in in an ARCHIVE run are specified.
More than one JOBVAR statement can be specified. Each JOBVAR statement may contain a number of NAME and EXCEPT operands, but only one EXCEPT operand per NAME operand.
The handling of the JOBVAR statement and the NAME operand is analogous to what was described for the FILES statement in the section "Parallel and serial processing".
J[OBVAR] |
[ { [NA[ME] = [ { jv-name / (jv-name[,THRU = jv-name],...) / (jv-name[,RENAME = jv-name],...)] ] |
NAME =
The job variables to be restored are specified in the NAME operand.
The possible specifications are dependent on the value of the CATID parameter.
The following applies to runs with
PARAM CATID=NO
:A catalog ID may be specified only in the RENAME operand. Otherwise, the statement will be rejected.
If the NAME operand is omitted:
All job variables are written into the default pubset of the user ID concerned.
If the user has the TSOS privilege, the job variables of all user IDs are written into their respective default pubsets.
The following applies to runs with
PARAM CATID=YES
:If no catalog ID is specified, the catalog ID of the default pubset of the relevant user ID is added to the job variable name.
If a specified user ID has no entry on the home pubset, this leads to an error. The job variables are not read in.
It is also possible to specify only the catalog ID (e.g.NAME=:c:
), in which case all job variables belonging to the user ID are restored which come from the catalog with the specified ID.If the NAME operand is omitted:
All job variables under the relevant user ID are restored to the pubset from which they were saved.
If the user has the TSOS privilege, the job variables of all user IDs are restored to the pubset from which they were saved.
NAME = jv-name
This may be a partially or fully qualified job variable name.
NAME = (jv-name,...)
If more than one job variable name is specified, they must be enclosed in parentheses.
NAME = (jv-name,THRU=jv-name)
The specification “THRU” enables a range of job variables to be defined. All the job variables lying alphabetically between the two names (both inclusive) are then addressed.
The job variables must be on one pubset in a run with PARAM CATID=YES
, i.e. the catalog IDs of the two job variables in jv-name,THRU=jv-name
must be the same.
NAME = (jv-name,RENAME=jv-name)
Using the RENAME operand, job variables can be renamed when they are read in in RESTORE and IMPORT runs. The name specified in RENAME replaces the original name (NAME=...
). If the name given is partially qualified, only the specified part is renamed. If no catalog ID :c:
is specified in the job variable names, the standard catalog ID is assumed.
A catalog ID (target pubset) may only be specified in RENAME in the case of PARAM CATID=NO
.
EXCEPT=jv-name / (jv-name,THRU=jv-name)
This specifies job variables or job variable ranges that are to be excluded from the list of job variables in the NAME operand.
Only one EXCEPT operand may be specified per NAME operand.
If the EXCEPT operand is not preceded by a NAME operand, it refers to all the job variables under the user ID or, in the case of the system administrator ID, to all the job variables in the system.
The same applies to use of the EXCEPT operand in the JOBVAR statement as was described for the EXCEPT operand in the FILES statement.
The rules governing the specification of job variable names in the NAME operand also apply to the EXCEPT operand.