Alan, Can you please specify the massive cluster overhead?In cache fusion, every time only up to 3 instances are involved, at least for often touched blocks.
(I currently doesn't know how the 'master' of an object is communicated) So if the interconnect-switch is designed well, it will scale. Martin Am 23.01.2010 um 22:59 schrieb Guillermo Alan Bort:
Actually, a 100node RAC would have a massive overhead in cluster administration and would almost certainly be slower than a smaller RAC environment...On the other hand... it might be faster than VMWare ;-) Alan.-
-- //www.freelists.org/webpage/oracle-l