Yes, and we've added automation to configure the zIIPs online when that message is issued. I've also gone in and defined the zIIPs online to those LPARs in the activation profile and removed the CF CPU,ON commands from COMMND00. I expect that those will remove the need for the automation (we set it up for the IWM064I message saying boost ended).
I'm sure the surprise we got was well known, but neither I (admittedly, this is my first install of a system with Boost - or zIIPs) nor our IBM Business Partner's guy were expecting it. I'm told that we set things up as we did back when we moved from a z10 to the z14; we had a couple of zIIPs "on loan" from our business partner to see if they'd improve our performance. (Spoiler: We're an Adabas and Natural shop. They did, dramatically.) We weren't sure we'd keep them, so we set it up that way so that if we decided to turn them off, we wouldn't fail to activate the system. At least this is the story I'm getting. Once we decided to keep the zIIPs, we simply never changed the configuration, since it was working. On Thu, Dec 18, 2025 at 8:24 AM Mike Schwab < [email protected]> wrote: > Is there a message saying the Boost is done, so the customer can > automate the final running CPU status? > > On Thu, Dec 18, 2025 at 8:14 AM Peter Relson > <[email protected]> wrote: > > > > This "surprise" is well understood and documented. The caveat of not > configuring processors online or offline while a boost is underway (in this > case the boost started before the config commands were processed) factors > into this. So does PRESCPU. > > > > The good news is that there is hope down the road that this scenario > will be improved.It requires both software and firmware support to > accomplish. The support needs are understood.Part of it has to do with > treating a config online for a zIIP during a boost as removing the zIIP > from being considered a transient/boost zIIP even though the zIIP was > brought online for the boost as a transient/boost zIIP (and, since the > consideration is removed, not taking that zIIP offline when the boost ends, > as is done for transient/boost zIIPs). > > > > When/if that will actually happen, I have no idea. > > > > Peter Relson > > > > ---------------------------------------------------------------------- > > 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 > -- Jay Maynard ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
