I believe oVirt draws the line at Nehalem, which contained important improvements to VM
performance like extended page tables. Your Core 2 based Xeon is below that line and
you'd have to change the code to make it work.
Ultimately oVirt is just using KVM, so if KVM works, oVirt can be made to work, too, and
KVM still supports much older CPUs.
I've faced similar issues when launching oVirt on Atoms, which are also considered
below that line, when in fact they support all Nehalem features. The problem was that
oVirt sets the basic CPU above the line, when it creates the self-hosted virtualized
management engine, which then fails to start because the CPU is below the line. By that
time the initial setup VM has already done its work, so it's a bit of a nasty surprise
and difficult to detect...
I got around the issiue by using a more modern CPU for the initial setup of my 3 node HCI
clusters and then downgraded the CPU baseline afterwards. But in theory you could just
find the code that sets the CPU type and change it, there is a good chance it's hidden
away in some Ansible or Python script.
Of course switching systems mid-flight comes with all kinds of other issues, but when
you're bent on bending the basic requirements the developers have used for their code,
you need to do the extra work.