MVSFORUMS.com

Cogito-Ergo-Sum · Posted: Sat Dec 28, 2002 1:27 am Post subject:

Kolusu,
Actually, I must apologise for asking a stupid question. A second later I posted my previous post, I realised the answer. Unforutnately, I could not edit my post then. (Otherwise, I would have missed my last bus to home!)

Thanks anyway, Kolusu!
_________________
ALL opinions are welcome.

Debugging tip:
When you have eliminated all which is impossible, then whatever remains, however improbable, must be the truth.
-- Sherlock Holmes.

rasprasads · Posted: Fri Jan 10, 2003 1:58 am Post subject:

I had a job to allocate a Sequential dataset with following code:

Cogito-Ergo-Sum · Posted: Fri Jan 10, 2003 2:21 am Post subject:

Rasprasads,
I do not get this.

Premkumar · Posted: Fri Jan 10, 2003 3:26 am Post subject:

rasprasads · Posted: Fri Jan 10, 2003 3:28 am Post subject:

Corgito,
I could not allocate the dataset in a single volume as there were large no. of records.I allocated a multi volume dataset.Would that be a reason for this?
(The dataset is in disk not in tape...)

If so could you throw some light on multi volume datasets allocation space?
_________________
Rasprasad S

Premkumar · Posted: Fri Jan 10, 2003 3:45 am Post subject:

rasprasads,

Are you sure that your output file is created in DASD and not in Tape?
An utility we have says 32670 is the efficient block size for a FB605 file in tape.

rasprasads · Posted: Fri Jan 10, 2003 3:52 am Post subject:

Premkumar,
As i said earlier the datset is in disk and not in tape.But it has been allocated as a multi volume dataset...
_________________
Rasprasad S

kolusu · Posted: Fri Jan 10, 2003 6:28 am Post subject:

rasprasad,

I guess your shop does not have SDB parm turned on. So even though you have coded blksize 0 it will not calculate the optimum block size.

I am also guessing that your input file is on tape and it's blocksize is 32760.

To copy and get the system determined blocksize use the following sysin card.

rasprasads · Posted: Fri Jan 10, 2003 6:53 am Post subject:

Kolusu,
Could you please look into this :

SYSIN :
OPTION SDB=ON
*
SORT FIELDS=COPY
OPTION STATEMENT : SYNTAX ERROR

I have coded the SYSIN as follows:

//SYSIN DD *
OPTION SDB=ON
SORT FIELDS=COPY

I am not aware of what SDB is!Can you please explain this...
_________________
Rasprasad S

kolusu · Posted: Fri Jan 10, 2003 7:08 am Post subject:

Rasprasad,

If you have DFSORT at your shop then specify

rasprasads · Posted: Fri Jan 10, 2003 8:34 am Post subject:

Kolusu,
Thanks for your solution.It gave the result.

But when i executed the same job once again with a different i/p file and mentioning a DATACLAS parameter for the Output dataset, i got the following message and The Dataset was allocated a blocksize of 32670 and gave the message:

$ORTPARM : SDB=ON
SYSIN :
SORT FIELDS=COPY

ALTERNATE PARM USED

BSAM WAS USED FOR SORTIN
BSAM WAS USED FOR SORTOUT

I hope the reason may be due to BSAM. Is it correct?
Also for 4.5 million records the no of cylinders allocated were around 997.
How is this arrived at?
_________________
Rasprasad S

hebbs · Beginner Joined: 14 Aug 2004 Posts: 1 Topics: 0

Does this calculation apply to VB files as well?
_________________
Hebbs

Mervyn · Posted: Sat Aug 14, 2004 3:45 pm Post subject:

No. Generally, use half-track blocking for disk files, and 32760 for tape files.

System Determined Blocksize, as defined above, is easier still.
_________________
The day you stop learning the dinosaur becomes extinct

Nagaraj.krish · Beginner Joined: 12 Apr 2006 Posts: 1 Topics: 0

Suppose the Block size = 27500
Rec length = 500

Then No of records in a block - 55.

What will be the no of records in a Track.

In previous discussion, i infer it will be 110 records in a track.
But later it is said that there can be more than 2 records /block/ track.

How to arrive the correct number atleast appx....

kolusu · Posted: Wed Apr 12, 2006 9:40 am Post subject:

Nagaraj.krish,

You will have 110 records in a track. The contradiction about the no: of records is for a different blocksize.

Kolusu
_________________
Kolusu
www.linkedin.com/in/kolusu