You know, all things considered, I think the simplest solution for **my** problem is to add a little delay after each SUBMIT. I know it's not elegant, and I know it's not bulletproof, but this is a development test script, not a production job. The total run-time for the 70 jobs is probably two hours (the three hour total includes think and edit time). If I add a 5-second delay after each SUBMIT, it will probably solve the problem, and only take six minutes in total, which will overlap, of course, with the two hours. I can still SUBMIT a heck of lot faster than I can knock off the jobs in batch.
It turns out I can recognize true collisions because the ALLOCATE OLD fails, presumably because the converter has the JCLLIB locked up DISP=SHR. (I would not recognize a situation where the ALLOCATE beat the converter to the punch.) If five seconds works, I may try two or three. If five seconds is not enough, I'll try ten. As I said, it's not elegant, but I think it's the best solution to my real-world problem. I've got more important fish to fry than finding the most elegant way to submit 70 regression tests. Charles -----Original Message----- From: IBM Mainframe Discussion List [mailto:[EMAIL PROTECTED] On Behalf Of Paul Gilmartin Sent: Tuesday, October 17, 2006 7:31 AM To: [email protected] Subject: Re: Collisions between converter(?) and TSO? In a recent note, Charles Mills said: > Date: Tue, 17 Oct 2006 07:07:13 -0700 > > I guess the negative would be the large time investment I have now in a > JCL-driven process, and the fact that it's a batch product so JCL is the > environment in which it is most likely to run at a customer's. > Rexx runs fine in batch, under either IRXJCL or IKJEFT01. And you can use ISP^H^H^HSDSF to browse the output of your EXEC. ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
