I am not sure if there was an easier way, but what I ended up doing worked pretty well.

* Staring with a working 8.0.0.GA.5434 64 FOSS edition install on Centos 6.3 in an Ubuntu 12.04 Host (KVM)
* Clean Ubuntu 12.04 64bit install
* cloned /etc/hosts to new host
* Installed zimbra on the new Ubuntu guest
* moved /opt/zimbra /opt/zimbra.clean
* I rsync'd the working /opt/zimbra to the new ubuntu 12.04 guest /opt/zimbra
* centos: stopped zimbra
* rsynced one more time for changes
* tried a start and it failed -- which I thought it would since the binaries are wrong.
* for the hell of it, I ran ./install.sh script again on the ubuntu host... surprise. It noticed zimbra was already installed and asked if I wanted to upgrade v8 to v8? YES
* after the upgrade, zimbra started and worked perfectly.

Looking back, I am guessing the bettter way to do this would be to just rsync /opt/zimbra from the working host to the new one and then run ./install.sh. I am just not sure if the install.sh script asked to upgrade due to the the fact the packages were installed or if the /opt/zimbra dir existed.. or both. I just documented my steps exactly since they proved true.

Note, for the migration I used the exact same hostname/IP address for the new host. I am running a virt environment (KVM) so I just paused the working host, brought up the ints on the new host and started zimbra.

Reason I migrated? I thought my Centos 6 guest was a bit slow (services starting up) and wanted to verify if it was my virt or zimbra... and why not just to test. After all of this, I have a working ubuntu host - but the services start up just as slow. So now I know, it's not related to the OS. Slow is also relative. I don't know what to expect time frame for startup / shutdown of zimbra services. All I know, is it takes my virt from booting in under 10 seconds to 60+ give or take.