Hi,
I can't speak for RedHat, but the reasons at the time appeared to
have been mainly technical, with the main reason being the fact
that Xen support was not upstreamed into the Linux Kernel at the
time (hardware, porting costs, etc. all follow from this). Whether
the size of codebase was a reason is actually debatable: it
depends on how you define KVM and Xen and measure size (e.g. do
the Xen codebase contains testcode and toolstacks, the KVM
codebase doesn't - KVM user components depend heavily on QEMU,
which is quite a sizeable codebase - etc.). Of course since then,
RedHat has heavily invested in KVM. The primary technical reasons
why RedHat may have done this, by now have gone away.
Besides building Xen from source, increasing momentum to bring Xen
back into the RedHat world is actually pioneered by a number of
Xen and CentOS community members. For example, see
https://fosdem.org/2013/schedule/event/xen_centos6/ ... I will
leave things at this, as I don't want to take away their thunder
in the coming weeks when this project starts to become more
public. I would expect that there also will be follow-up projects
for CentOS 7, etc.
Lars
On 29/01/2013 09:29, tech mailinglists wrote: