Like parameter with GDG

kranthikumarb · Posted: Wed Feb 18, 2009 10:52 am

Hi,
I am trying to use the following JCL and procedure to run a simple program which writes into two files. It is throwing me the following error

dick scherrer · Posted: Wed Feb 18, 2009 11:16 am

Hello,

Are these files sms managed? I believe this option only works with sms managed data.

gcicchet · Posted: Wed Feb 18, 2009 12:47 pm

Hi,

I think the problem is you cannot use LIKE using the ralative GDG, if you specify the fully qualified name you will overcome the syntax error.

Gerry

expat · Posted: Wed Feb 18, 2009 1:03 pm

Or maybe it is because there is no value for GENIN being set but &GENIN is used in your PROC.

kranthikumarb · Posted: Wed Feb 18, 2009 1:20 pm

That was just a copy paste error. I took the wrong version of the JCL Please ignore GENOUT/GENIN problem. I still have the same error even after changing it to GENIN.

dick scherrer · Posted: Wed Feb 18, 2009 8:18 pm

Hello,

kranthikumarb · Posted: Thu Feb 19, 2009 9:22 am

Hi Dick sherrer,

As suggested , I started from the begnining of the JCL i.e. step by step and i found that it is working fine if i use the latest version of the gdg manually.
But if i replace G0002V00 with (0). It is throwing me error. Is there any way to point to the latest version of the GDG?

dick scherrer · Posted: Thu Feb 19, 2009 9:39 am

Hello,

kranthikumarb · Posted: Thu Feb 19, 2009 9:50 am

Well, I tried using (0) only.

I tried using (+1) for creating a new generation dataset and for referring to the latest version of previous file, i used (0) in LIKE parameter. I dont see any problem with this. Correct me if i am wrong.

Also, we are using JES3. That means all the datasets are allocated before the job is scheduled. So a new version of the dataset OUT1 must be created already before the job is scheduled.

dick scherrer · Posted: Thu Feb 19, 2009 10:23 am

Hello,

kranthikumarb · Posted: Thu Feb 19, 2009 10:45 am

Yes, I used sequential files instead of GDGs.

vvrahul · Posted: Thu Feb 19, 2009 3:26 pm

Hi,

I would like to add to Dick. When new GDG version are being allocated, the catalog happens only when the job finishes.

//STEP01 EXEC PGM=IEFBR14
//SYSOUT DD SYSOUT=*
//SYSPRINT DD SYSOUT=*
//DD01 DD DSN=&ID.FILE.GDG(+1),
// DISP=(NEW,CATLG,DELETE),
// DCB=(LRECL=80,BLKSIZE=0,RECFM=FB),
// SPACE=(TRK,(1,5),RLSE),
// UNIT=SYSDA
//DD02 DD DSN=&ID.FILE.GDG(+1),
// DISP=(NEW,CATLG,DELETE),
// DCB=(LRECL=80,BLKSIZE=0,RECFM=FB),
// SPACE=(TRK,(1,5),RLSE),
// UNIT=SYSDA

Would allocate and catalog only one verion of the GDG.

If we do need two version one should use:

//STEP01 EXEC PGM=IEFBR14
//SYSOUT DD SYSOUT=*
//SYSPRINT DD SYSOUT=*
//DD01 DD DSN=&ID.FILE.GDG(+1),
// DISP=(NEW,CATLG,DELETE),
// DCB=(LRECL=80,BLKSIZE=0,RECFM=FB),
// SPACE=(TRK,(1,5),RLSE),
// UNIT=SYSDA
//DD02 DD DSN=&ID.FILE.GDG(+2),
// DISP=(NEW,CATLG,DELETE),
// DCB=(LRECL=80,BLKSIZE=0,RECFM=FB),
// SPACE=(TRK,(1,5),RLSE),
// UNIT=SYSDA

Similarly I used the Like function and it worked fine for me:

//STEP01 EXEC PGM=IEFBR14
//SYSOUT DD SYSOUT=*
//SYSPRINT DD SYSOUT=*
//DD01 DD DSN=&ID.FILE.GDG(+1),
// DISP=(NEW,CATLG,DELETE),
// DCB=(LRECL=80,BLKSIZE=0,RECFM=FB),
// SPACE=(TRK,(1,5),RLSE),
// UNIT=SYSDA
//DD02 DD DSN=&ID.FILE.GDG(+2),
// DISP=(NEW,CATLG,DELETE),
// LIKE=&ID.FILE.GDG.G0001V00

Thanks,

Rahul.

Peter Poole · New User Joined: 07 Jan 2009 Posts: 50 Location: Scotland

Correct. (+n) relative GDGs are used to work out what the explicit G**V* suffix will be, based on a look at the catalog when the job starts. (Though in the example above the OP is creating +1 of two different datasets)

If you can't get like to work with relative GDG and don't want the hassle of working out what the explicit DSN is, you could always dodge the bullet by setting up a spurious non-GDG file as ID..&QUAL1..MODEL and using that in the LIKE statement.

Cheers.

ykishor · New User Joined: 11 Aug 2007 Posts: 24 Location: my pc