Re: Xen, Squeeze, and Beyond
On Fri, Feb 26, 2010 at 11:18:41AM +0100, Goswin von Brederlow wrote:
> >> "Xen - Provides para-virtualization and full-virtualization. Mostly used
> >> on servers. Will be abandoned after squeeze."
> > I think that the problem here is that Xen isn't mainstream in the
> > kernel. It takes a long time for a Xen-ified kernel to come out and any
> > distribution supporting it has to carry a heavy patch burden. Xen
> > doesn't keep anywhere current in terms of kernel - if we release Squeeze
> > this year with kernel 2.6.3*, Debian will have to maintain all the patches
> > / "forward port" them to 2.6.32 or 2.6.33 as was done with 2.6.2*.
> I think we can all agree that the old style xen patches from 2.6.18 and
> forward ported to newer kernels in lenny are unmaintainable.
> But the pv-ops xen kernel is shaping up well and that is what Bastian
> Banks is working on. They have a proper upstream and follow the latest
> vanilla kernel well enough. According to the wiki the plan is to have
> pv-ops merge into vanilla with 2.6.34.
Let's not concentrate too much on having dom0 support in mainline, because
that is not a panacea - we (Debian) just need a stable Xen patch that tracks
the current stabler-mainline branch or whatever we're targeting for our
stable release, and has some sort of a forseeable future maintenance path.
We have shipped .26 patches in lenny and they turned out to be really buggy
in some cases (domU .26 kernels with vcpus >= 1 can get random freezes on
.26 dom0), and it's not getting fixed because that patch branch is EOL'd.
That's worse of a problem for users than some theoretical later abandoning.
Since we're concentrating on .32 now, the paravirt_ops branch looks good,
because Xen upstream agreed to form their own .32 stable branch of that.
So whether they succeed in a mainline merge for .34 or .35 or .36 or even
later, that's irrelevant, we will still have support for the squeeze kernel.
2. That which causes joy or happiness.