On Sat, 7 Apr 2018 20:33:01 +0000, Nims,Alva John (Al) wrote:
>You are correct that the input data sets cannot be concatenated, but I might
>give an update to the solution in that you can specify multiple input DD's on
>the copy statement.
>From the examples for merging 4 data sets into one single one, but with an
>addition of the select statement:
> //SYSPRINT DD SYSOUT=A
> //IN1 DD DSNAME=DATASET1,UNIT=disk,VOL=SER=111112,
> // DISP=SHR
> //IN5 DD DSNAME=DATASET5,UNIT=disk,VOL=SER=111114,
> // DISP=OLD
> //OUT2 DD DSNAME=DATASET2,UNIT=disk,VOL=SER=111115,
> // DISP=(OLD,KEEP)
> //IN6 DD DSNAME=DATASET6,UNIT=disk,VOL=SER=111117,
> // DISP=(OLD,DELETE)
> //SYSUT3 DD UNIT=SYSDA,SPACE=(TRK,(1))
> //SYSIN DD *
> COPYOPER COPY OUTDD=OUT2
> INDD=IN1
> INDD=IN6
> INDD=IN5
> SELECT MEMBER=(ABC)
> /*
>
>This way the first data set that the member is found it is used and no error
>will be generated.
>
I RTFM (DFSMSdfp Utilities) which confirms what you say. Curious, I read
further.
It appears, but only implicitly, that IEBCOPY supports PDSU to PDSU copy with
member
selection and exclusion:
IEBCOPY can copy a PDSU to a PDSU directly without the need to reload it to
a PDS
and then unload the PDS to create the new PDSU. If a selective copy is not
required
then it will be faster to use IEBGENER to copy the PDSU to a new PDSU.
This suggests (exception that proves the rule?) that selective copy from PDSU to
PDSU is permitted with some performance penalty. An example would be
informative.
Also:
To unload more than one partitioned data set to the same tape volume in
one execution of IEBCOPY, multiple copy operations must be used and multiple
sequential data sets must be allocated to successive files on the tape.
I suspect the mention of tape is archaic. Does the same restriction apply if
the
output (sequential) PDSU is on DASD?
Can a single reload/copy operation succeed with multple input data sets where
some are PDS(E) and others are (sequential) PDSU? An example would be
informative.
The discussion of the order of operations seems almost superflous. Can this
even
be discerned by the user unless the output data set is sequential in which case
only
a single input data set is allowed? Then, I'd expect members to be unloaded in
directory (alphabetical) order.
What happens if some member names in input data sets contain prohibited
characters?
I'd expect they can't appear in SELECT or EXCLUDE statements.
-- gil
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
