I agree with Dave, that there can be hidden problems.
However, if you still want to purge RC=0 jobs, refer to the JES2 Initialization
and Tuning Reference.
The JOBCLASS/STCCLASS/TSUCLASS parmlib statement has several parameters which
can be used:
CONDPURG=Yes|No
Specifies whether (YES) or not (NO) system data sets (such as JESMSG and
SYSMSG)
in this job class are to be conditionally purged.
This parameter is not allowed on the JOBCLASS(v) initialization statement.
Modification: $T JOBCLASS or cold start.
OUTDisp=([normal] [,abnormal])
Specifies the default output disposition for JES data sets created for jobs in
this class, depending on whether or not the job abended.
normal
Specifies the output disposition if the job does not abend.
abnormal
Specifies the output disposition if the job abends.
Valid output dispositions are:
Hold
Hold the output. JES2 does not process the output until you either
change the disposition to WRITE or KEEP, or release the output. When
the output is released, the disposition changes to WRITE.
Note: This specification is honored when the output reaches the
destination node.
Keep
Process the output and then keep a copy of it on spool. After
processing, the disposition of this output becomes LEAVE.
Leave
JES2 does not process the output until you change the disposition to
WRITE or KEEP, or release the output. When the output is released,
the disposition changes to KEEP.
Purge
Purge the output immediately.
Write
Process the output then purge it.
Only code OUTDISP= on JOBCLASS when you require that system data sets
have a specific output disposition regardless of their message class
(MSGCLASS=) specification. If you have no such requirement, either code
OUTDISP=(), OUTDISP=(,) , or leave it off the JOBCLASS statement
completely.
If you do not code OUTDisp=, JES2 assigns the value on the OUTCLASS
statement for JES2 member data sets in this class. See z/OS JES2
Initialization
and Tuning Guide for more information about specifying output disposition and
how JES2 determines output disposition defaults.
Modification: $T JOBCLASS operator command or cold start
Don
> -----Original Message-----
> From: IBM Mainframe Discussion List [mailto:IBM-MAIN@LISTSERV.UA.EDU]
> On Behalf Of David Devine
> Sent: Thursday, February 21, 2013 10:51 AM
> To: IBM-MAIN@LISTSERV.UA.EDU
> Subject: Re: want to purge all my jobs ended with a RC=0 automatically
>
> Hi Yann,
> there is a big flaw in this, in that the fact that the job has a rc of
> zero doesnt always mean that it has worked sucessfully as designed.
> I think you would be better off using RMDS (or a similar product) to
> offload production joblogs from your spool to disk and have sms assign
> a mgmtclas to delete them after x days.
>
> This would get them off your spool and still allow the opportunity to
> check them if needed.
> There would be an overhead in disk usage but it's probably worth it.
>
> For everything else there's a load of options for the jes commands to
> delete after hours/days.
> POJQ allows job masking.
>
> There may well be a Jes exit which will achieve what you are looking
> for, but not really my field.
> A trawl through the jes manuals should identify posibilities.
>
> regards
> Dave
>
> ***********************************************************************
> ************************
>
> Hi all,
>
> For some volume reasons, we want to automatically purge jobs that ended
> with a RC=0.
> But we don't want to do it, for CPU and Spool reasons, with a JES2
> command. We want to do it automatically.
> I'm sure that somebody already had the same need, no ?
>
> Any Idea ?
>
> Thank in advance ...
>
> Yann
>
>
>
>
>
> ----------------------------------------------------------------------
> For IBM-MAIN subscribe / signoff / archive access instructions,
> send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
