Thanks Jim that is clear.  Why not include V=F support for lpar mode
systems?  This will then need to be without i/o assist, but given the z/arch
implementation v=f support could be implemted rather easy.  This could be
very similar to a v=v guest, but would use a z/arch real space, with offset
and extent defined in the state descriptor.  This way there will notr be a 2
level virtual storage.  2 levels of virtual storage used to give a
performance panelty when running aix/esa under vm/xa using veoctor
processors.  Now this is a long time ago, but there are probably still
workloads around that do have a performance issue with multiple levels of
virtual storage (tpf?)

Jan Jaeger.

From: Jim Elliott <[EMAIL PROTECTED]>
Reply-To: Linux on 390 Port <[EMAIL PROTECTED]>
To: [EMAIL PROTECTED]
Subject: Re: OT: Intel gets virtualization clue?
Date: Wed, 8 Oct 2003 21:32:58 EDT

> What is the status of multiple preferred guests on z990 machines?

> iirc the z990 cannot run in basic mode, which was always a prereq
> for multiple preferred guests. When running under PR/SM the MHPGF
> would be taken by PR/SM. Running multiple preferred guests will
> require something like 2nd level zones, unless PR/SM has changed
> such that it no longer uses the MHPGF.

Jan:

Unfortunately, you loose support for V=F guests on the z990. With the
complexity of the I/O subsystem (two logical channel sets, 512
channels) LPAR mode became mandatory to manage the environment. If you
have images that really require the level of performance provided by
V=R/V=F support, you should probably run those directly in an LPAR.
With 30 LPARs, you have support for more "native" systems than before.

Jim

_________________________________________________________________ Chatten met je online vrienden via MSN Messenger. http://messenger.msn.nl/

Reply via email to