I believe this is an effort hat is easily handled by a scheduling software.
However, I had to do something like this, and I wrote all my SYSOUT from the JOB to a dataset, then the last step would parse it to see if the next job could be submitted. RC=04 was not always a good thing. And the process had to account for new or renamed members. Or someone adding something to the dataset not expected. Lizette > -----Original Message----- > From: IBM Mainframe Discussion List <IBM-MAIN@LISTSERV.UA.EDU> On Behalf Of > Daniel S. Dalby > Sent: Saturday, April 21, 2018 11:43 AM > To: IBM-MAIN@LISTSERV.UA.EDU > Subject: Re: Mass submit jcl > > Hi Lizette, > > My thought was that the poster could submit the members in alphabetic > sequence. > As the job was submitted it would tack on an additional step to the end to go > through the submission library and submit the NEXT member IF the return code > was zero. Each step as they got submitted would add the additional step to > indicate where it came from and what member the current JCL came from. The > last step would check if there is another member after this one in the > library and submit it along with the additional step ... and so on until a > non-zero return code is encountered or the last member is submitted. > > As you suggested, a scheduler would be the best method but the original > poster didn't say if they had one or not. > I don't know if the new IBM SCHEDULE statements allow for return code > checking in a submission chain. > > Also, by sequentially submitting the jobs in alphabetical sequence and not > submitting the next until the previous job has completed successfully, you > won't get jobs running out of order. > > DanD > > "Lizette Koehler" <stars...@mindspring.com> wrote in message > news:<0bd601d3d993$0e7d02d0$2b770870$@mindspring.com>... > > Do you have a job scheduling software? If not, then submitting in a > specific > > sequence will be tricky > > > > Are you at z/OS V2.3? If so, JES2 now has some scheduling control > > cards > that > > will allow you to do simple scheduling functions. > > > > Check on www.cbttape.org for scheduling functions. > > > > > > If the jobnames are different, the when they hit the input queue they > might run > > out of order. > > > > If you provide more specific details, we might provide better answers > > > > > > Lizette > > > > > > > -----Original Message----- > > > From: IBM Mainframe Discussion List <IBM-MAIN@LISTSERV.UA.EDU> On > > > Behalf > Of > > > Daniel S. Dalby > > > Sent: Saturday, April 21, 2018 9:36 AM > > > To: IBM-MAIN@LISTSERV.UA.EDU > > > Subject: Re: Mass submit jcl > > > > > > <ss33ss...@gmail.com> wrote in message > > > news:<6462565f-68ca-445b-9226-b25257e1e...@googlegroups.com>... > > > > Is there any methods for submitting all the members in a pds in > > > > order and > > > do return code checking at the job level so that job b is depends on > joba > > > rc=0 etc > > > > > > --- > > > > > > You'd get better response if you submitted to the mailing list > > > rather > than > > > posting to the newsgroup. Fortunately some of us read from the > newsgroup > > > which contains both. > > > > > > Here's a challenge to yourself or anyone interested in playing for a > > > few hours. I'm busy this weekend otherwise I'd give it a try. > > > > > > Create a procedure "SUBNEXT" where you pass it a PDS/PDSE library > > > name > and > > > the current member. Write it in whatever language you are most > comfortable. > > > A CLIST or REXX could be used as the PROC could contain "//substep > > > EXEC PGM=IKJEFT1A,PARM='%SUBNEXT &dsname &member'". > > > > > > The SUBNEXT program would allocate the library, create a member > > > list, > find > > > the current member and check if another member follows, if so submit > that > > > member. > > > The routine could dynamically allocate the INTRDR and using an > > > ACB/RPL > submit > > > via PUT the JCL onto the internal reader and at the END of that > > > input > member > > > add ... > > > "//LASTSTEP EXEC SUBNEXT,DSNAME=dsname,MEMBER=mem,COND=(0,NE)". > > > > > > When using an ACB/RPL to write to the internal reader after you > > > issue > ENDREQ > > > the RPLRBAR field will contain the 8 character job id, which can be > > > used > in > > > any messages if you wish. > > > It would be nice to have another step following the "SUBNEXT" step > > > that issues a write to operator if SUBNEXT was NOT executed. I > > > guess that > could > > > be up to the user if they wanted that functionality. > > > > > > SUBNEXT would NOT be a mass submission of the PDS/PDSE but rather > sequential > > > submission of the members within that library. By using COND= > submission > > > would STOP whenever a job does NOT work as expected (in this example > > > the > job > > > does not end with a return code of zero. > > > > > > If nobody attempts this in the next few days, I'll give it a shot. > > > I > could > > > find this useful for submitting a TEST job streams. > > > > > > Please post to the LIST if you give it a shot and successfully > > > complete > this > > > tool. I would like a copy. > > > > > > Thanks, > > > DanD > > ---------------------------------------------------------------------- > 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
