View previous topic :: View next topic
|
Author |
Message |
senjay
Active User
Joined: 10 May 2007 Posts: 147 Location: India
|
|
|
|
Hi,
I have used DISP=(NEW,PASS,DELETE) for a file and it has been passed but not deleted at the end of the job (Successful execution). Could any one explain why? |
|
Back to top |
|
|
CICS Guy
Senior Member
Joined: 18 Jul 2007 Posts: 2146 Location: At my coffee table
|
|
|
|
Was there any other references to that DSName (before or) after the one you are showing? |
|
Back to top |
|
|
senjay
Active User
Joined: 10 May 2007 Posts: 147 Location: India
|
|
|
|
Hi,
I am creating this Dataset using DFSORT in the first step
DSN=MY.PASS.DATASET
DISP=(NEW,PASS,DELETE),
UNIT=SYSDA,
SPACE=(CYL,(500,500),RLSE),
DCB=(LRECL=3000,RECFM=FB,BLKSIZE=0)
and then using it in the next step (i.e, Program execution Step). There are no more steps. Please get back if you need more info |
|
Back to top |
|
|
dick scherrer
Moderator Emeritus
Joined: 23 Nov 2006 Posts: 19244 Location: Inside the Matrix
|
|
|
|
Hello,
You need to post the "next step" as well. |
|
Back to top |
|
|
senjay
Active User
Joined: 10 May 2007 Posts: 147 Location: India
|
|
|
|
Hi Dick,
please see my execution step below
Execution step:
//S030 EXEC PGM=PROGRAM,COND=(0,LT)
//STEPLIB DD DSN=MY.LOADLIB,DISP=SHR
//EIN1 DD DSN=MY.PASS.DATASET,
// DISP=SHR
//RMCUST DD DSN=INPUT.FILE.FROM.ANOTHER.JOB,
// DISP=SHR
//AUS1 DD DSN=MY.OUTPUT.FILE,
// DISP=(NEW,CATLG,DELETE),
// UNIT=(3390,12),
// SPACE=(CYL,(500,800),RLSE),
// DCB=(RECFM=FB,LRECL=570,BLKSIZE=0) |
|
Back to top |
|
|
dick scherrer
Moderator Emeritus
Joined: 23 Nov 2006 Posts: 19244 Location: Inside the Matrix
|
|
|
|
Hello Senthil,
Look at the 3rd sysout and see how JES disposed of the file.
This
Quote: |
//EIN1 DD DSN=MY.PASS.DATASET,
// DISP=SHR |
does not make sense to me. If the file is not needed, the disp should probable be (old,delete,delete). If the file is to be preserved so that this step could be restarted (rather than running from the beginning), the disp should be (old,delete,catlg). I can think of no reason to use shr in a newly created, passed dataset.
If this is truly a "throw-away" dataset, you could use a dsn of &&MYPASS. |
|
Back to top |
|
|
senjay
Active User
Joined: 10 May 2007 Posts: 147 Location: India
|
|
|
|
Hi Dick,
Thanks for the explanation. I understood. But i still have one doubt. The above job has run 5 times and it is throwing the problem now only. Still wonder if this depends on some other factor also? |
|
Back to top |
|
|
dick scherrer
Moderator Emeritus
Joined: 23 Nov 2006 Posts: 19244 Location: Inside the Matrix
|
|
|
|
Hello Senthil,
Quote: |
Still wonder if this depends on some other factor also?
|
Probably. The way the jcl is/was(?) running could cause problems eventually - maybe just now.
It is possible that there are still multiple other copies of this file still on dasd.
There may have been some change to the SMS rules.
You may want to talk with the storage admin prople and make sure that this dataset is not "just lying around" on some other dasd volumes.
Just because a job runs to eoj does not mean that some silent problems are not happening. Many organizations have jcl standards that dis-allow such things as well as having a technical review of the jcl before it may be promoted to production. |
|
Back to top |
|
|
senjay
Active User
Joined: 10 May 2007 Posts: 147 Location: India
|
|
|
|
Hi Dick,
I understood. Thanks much for your explanations. |
|
Back to top |
|
|
CICS Guy
Senior Member
Joined: 18 Jul 2007 Posts: 2146 Location: At my coffee table
|
|
|
|
Kind of make me think that the original problem and the following discovery that it was suddenly not being deleted is because the sixth time it was run the allocation could not find a volume where it had not already been created previously...... |
|
Back to top |
|
|
Anuj Dhawan
Superior Member
Joined: 22 Apr 2006 Posts: 6250 Location: Mumbai, India
|
|
|
|
Hi,
Little more observaion for this
Quote: |
The above job has run 5 times and it is throwing the problem now only. Still wonder if this depends on some other factor also? |
Did COND in this
Code: |
//S030 EXEC PGM=PROGRAM,COND=(0,LT) |
allowed to run this step in 'those' 5 runs ? |
|
Back to top |
|
|
senjay
Active User
Joined: 10 May 2007 Posts: 147 Location: India
|
|
|
|
Yes. All the 5 times, it allowed to run till the end. |
|
Back to top |
|
|
dick scherrer
Moderator Emeritus
Joined: 23 Nov 2006 Posts: 19244 Location: Inside the Matrix
|
|
|
|
Hello,
Quote: |
Did COND in this Code:
Code: |
//S030 EXEC PGM=PROGRAM,COND=(0,LT) |
allowed to run this step in 'those' 5 runs ? |
Quote: |
Yes. All the 5 times, it allowed to run till the end. |
Yup, i would expect this. Nothing is "wrong", so the job runs "successfully". Not really, but no error is thrown.
You may want to review other jobs that pass datasets that are not properly disposed of that may be waiting to cause some kind of surprise later.
Quote: |
because the sixth time it was run the allocation could not find a volume where it had not already been created previously...... |
While at a client site for a few days for something completely unrelated, we were interrupted because some critical job "went down" and needed attention. Seems like a situation similar to this had occurred and the problem file had been propagated across every volume in the center (except the spool, sysres, and other system support volumes). This was when everything was done by the unit parameter (befoer sms) and almost everything in the center was visable to SYSDA. . .
Quote: |
I understood. Thanks much for your explanations. |
You're welcome |
|
Back to top |
|
|
expat
Global Moderator
Joined: 14 Mar 2007 Posts: 8797 Location: Welsh Wales
|
|
|
|
Quote: |
If this is truly a "throw-away" dataset, you could use a dsn of &&MY.PASS.DATASSET |
I always thought that && datasets could not be more than one qualifier. |
|
Back to top |
|
|
dick scherrer
Moderator Emeritus
Joined: 23 Nov 2006 Posts: 19244 Location: Inside the Matrix
|
|
|
|
Hi Expat,
Quote: |
I always thought that && datasets could not be more than one qualifier. |
Good catch! I got carried away with the copy/paste
The dsn above has been corrected.
d |
|
Back to top |
|
|
|