It is better for performance to do the TRANSWAP before acquiring a lot of storage, since that will reduce the number of fixed frames than need to be moved to
non-reconfigurable storage by the TRANSWAP. However, the order of the TRANSWAP relative to the DSPSERV CREATE should have no effect on the 0D5 abend, which would have occurred because the address space was swappable (and logically swapped) at the time of the storage access which got the abend. Jim Mulder z/OS Diagnosis, Design, Development, Test IBM Corp. Poughkeepsie NY > Although it's possible, I doubt we have a bug in our logic in this > area; this has run for more than 10 years at multiple sites on > multiple levels of z/OS. And we do not ever issue a TRANSWAP > OKSWAP; the dataspace is designed to be accessible until the job > ends. Rob Scott's suggestion that moving The SYSEVENT TRANSWAP > before the DSPSERV CREATE is worth trying especially if we can find > a way to reproduce the abend. > > If the abend recurs at this site (or anywhere else), I will surely > get the output of a D J and an SVCDUMP. > > Since no one suggested it, I guess it is not likely that the site > would have a monitor or exit that would change a non-swappable > address space to swappable. ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
