On Wed, Jul 23, 2008 at 12:43:25PM +0200, Stefan de Konink wrote:
Or you could say that libvirt is broken because it isn't able to
distribute the inactive domains across the network in a consistent way.
But I think we already had that discussion.
Yes that discussion had been raised quite a few time, and I think
it's good to try to clear things up a bit:
http://libvirt.org/intro.html
that page is probably one of the oldest in libvirt documentation,
and I set there what to me is the fundamentals of the project:
"we can define the goal of libvirt: to provide the lowest possible
generic and stable layer to manage domains on a node"
a few more Node specific extensions have been adeed to this original
picture like Network and Storage available for the Domains, but I think
this clearly defines the scope of libvirt, and the intent is not to manage
a cluster. This can be done on top, this is being done actually by many
projects. I'm sorry but I don't want to try to extend libvirt in that
direction, there is no reason to put the emphasis toward one specific
management tool or setup.
There are still quite a few things to be done to improve libvirt
within the goal of the project, in addition of the constant evolution of
hypervisors facilities. The promise we make to the users is that they
need not to worry (too much) about low level changes, and can build their
solutions on top. The other implicit promise which I'm fine making explicit
is that libvirt won't make their life harder by growing outside the
initial project goals and conflicting with their own way of actually
managing the resources.
Daniel
--
Red Hat Virtualization group
http://redhat.com/virtualization/
Daniel Veillard | virtualization library
http://libvirt.org/
veillard(a)redhat.com | libxml GNOME XML XSLT toolkit
http://xmlsoft.org/
http://veillard.com/ | Rpmfind RPM search engine
http://rpmfind.net/