On Wed, 23 Jun 2010 20:47:24 -0400, Lizette Koehler wrote:

>It is going to depend.
>
>You can write a REXX that uses the LM functions to delete a member.
>If you have PDSMAN with FASTCOPY you can ZERODIR.
>If you have SAS you can KILL a pds.
>
>And I am sure there are other ways.
>The requirement is that the dataset is enqueued with DISP=SHR and that the
>SCHEDULER does not try to load from it when you are emptying it.
>
Delete every member, using the technique of programmer's choice.
For performance work in reverse alphabetical order.

Compress with DISP=SHR.  Integrity concerns are left to the user.
More generally, SCHEDULER isn't necessarily the only victim.

For performance, nearly the best is STOW DCB,,I.  Nearly?  Overwriting
the first directory block and resetting DS1LSTAR is probably better.
But that's incompatible with PDSE.  But the OP didn't say PDSE,
after all.

Delete and re-create is not an option while the data set is allocated.

STOW DCB,,I won't handle Unix directories, nowadays otherwise
(partly) supported by BPAM.  Mounting via NFS and using "rm -r"
handles all of PDS, PDSE, and Unix directories.  I won't
recommend it, although I've surely used it when it was convenient.

Probably technically the best is STOW DCB,,I using any language
which supports it at runtime.

>> Does anyone know of  a way to delete all of the members of a PDS which is
>> allocated by job scheduler software?  We currently use an assembler program
>> that does a reset(I think), but management wants us to not use assembler.  I
>> know that IEHPROGM can delete a specific member, but the members change
>> from day to day.  I don't think you can wild card it.

Will the OP's scheduler tolerate the use of Unix directories (now
partly supported by BPAM) as its libraries?  These could remove
most of the ENQ bottleneck while providing LUW isolation similar
to that of PDSE.

-- gil

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Reply via email to