If you have a highly kneecapped CP, almost any workload on a zIIP would greatly help the CP. Unless the instruction sequences are very short.
On Wed, Jul 20, 2016 at 10:14 AM, Paul Gilmartin <[email protected]> wrote: > On Wed, 20 Jul 2016 14:47:49 +0800, Timothy Sipples wrote: > >>Different products mean different technical considerations and >>optimizations. It's just that simple. For DB2 Sort for z/OS, zIIP >>exploitation makes technical sense. For DFSORT -- except for exploiting >>zIIPs on behalf of DB2 utilities and in other ancillary ways -- it doesn't >>seem to make technical sense. Maybe in the future, as the technologies >>change and evolve, it will. >> > I find it hard to believe that optimization of revenue was not a > consideration. > Andy White's post following yours suggests that, and that the choice may > have been counterproductive. > > -- gil > > ---------------------------------------------------------------------- > For IBM-MAIN subscribe / signoff / archive access instructions, > send email to [email protected] with the message: INFO IBM-MAIN -- Mike A Schwab, Springfield IL USA Where do Forest Rangers go to get away from it all? ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
