On Thursday, 12/20/2001 at 01:56 EST, David Boyes <[EMAIL PROTECTED]> wrote:
> >the 2001-12-12 sections describing the 31-/64-bit OCO-modules > >have been updated to clarify on HiperSockets support being > >currently available in LPAR, only, and noting the prereq on > >future VM-service (PTFs) for HiperSockets support for > >Linux systems running as VM-guests. > > I'm curious. What makes a virtual hipersocket under z/VM and a LPAR-based > hipersocket different? Shouldn't the virtualization in z/VM erase that > difference, or is this one of those "need to know" things that the driver > uses to turn on some HW assist? They aren't supposed to be different, except that virtual HiperSockets don't have the same limits (number of LANs, number of subchannels) as real HiperSockets. (Well, there is another important difference: virtual HiperSockets are implemented entirely in software - no exploitation of HiperSockets hardware.) Regards, Alan Senior Software Engineer z/VM, TCP/IP, VIF, and Linux for zSeries Development, Endicott, NY Phone 607.752.6027 fax 607.752.1497 t/l 852
