> Ahhh...  but could you have picked something other than Canis Majoris
> (the "Dog Star")???  Solaris for x86 already gets snarky remarks as
> being pronounced "Slolaris"...

To paraphrase a recently bailed-out automobile industry, this isn't your
father's Solaris. Especially on z10s. 8-)

> Won't run on bare metal, eh?  Oh, well...  so it avoids SIOs, but,
then,
> SX/1100 went a long way to avoid having to do any real I/O of its own,
> too.

Given the intended target of all those pizza-box G2 and G3 SPARCs about
to meet their maker by death or obsolescence, it just doesn't make sense
to go to all the effort of handling the raw devices if you can get z/VM
to do it for you, and get all the other goodies that z/VM does wrt to
caching and system management. 

Maybe someone will convince me otherwise, but my position on Linux in
LPARs is pretty clear too -- the kinds of workloads that need LPAR (vs a
VM guest) are probably somewhat questionable on Z in the first place if
you need more than just data proximity. The z10 helps, but it's a bigger
question of infrastructure and data management. 

The other goal of the z/VM orientation of the OpenSolaris port was to
truly get the virtualization concept front and center. I don't WANT to
propagate the idea that you run this on separate physical machines;
everything should be in VIRTUAL machines. Period. This isn't a toy; it's
a real statement that virtual infrastructure is the way to build the
next generation of data center. This is part of our plan to make it so. 

There's a lot of work going on between SNA and the big vendors (and not
just IBM) to develop this idea. Watch the trade press for how it
develops. 

----------------------------------------------------------------------
For LINUX-390 subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390

Reply via email to