[Oneiric-Topic] Revisit Xen support

Raphaël Pinson raphink at gmail.com
Sat Apr 2 15:26:53 UTC 2011


On Sat, Apr 2, 2011 at 5:03 PM, Clint Byrum <clint at ubuntu.com> wrote:
> Noted. It sounds like Xen has a lot of inertia.

I'd say big companies have a lot of inertia. Once they have chosen to
invest in a technology, be it Xen or KVM, they build systems on this
technology for years, hire experts (and get their employees to become
experts), and they're not willing to reconsider their position a few
months after (or even years) after just because the trend has changed.
Xen was the state-of-the-art not so long ago, and it's already taken
these big companies long enough to use it. KVM is the medium-to-long
term future for many still.


> I feel like there is a lot of anecdotal evidence, but we shouldn't make
> our decisions just because somebody says KVM can't do this or Xen can
> do that.

I think many would agree that with current hardware and a brand new
DC, it's generally a better idea to invest in KVM than Xen. But if we
want Ubuntu to be an enterprise class OS for servers, that means we
need to provide the tools for people to be conservative if they want
to, and not to have to rethink the technology they've been
successfully using for years. So, encouring the use of KVM, I'm all
for it ; supporting the people who still need to use Xen and would
like to do so with Ubuntu, that's also a major point if you want
Ubuntu to be a reference in DCs.

Nowadays, who would use inetd to implement a network service if they
have xinetd at hand? I guess not many people. But we're not going to
remove inetd from Ubuntu, because a lot of old services still use it
and it's a standard per se, whether we use it for internal Ubuntu
development or not.


Raphaël




More information about the ubuntu-server mailing list