Open Stack

Hi,

Just wondering if this is possible:
We have some hypervisors which are dedicated for running database instances.
I'd like to use a ram_allocation_ratio of 1.0 for these but 1.5 for the other hypervisors.

Is this something that can be set per machine or is this global?
The documentation says:
"This filter can be configured to allow a fixed amount of RAM overcommitment by using the ram_allocation_ratio configuration option in nova.conf."
But is this setting only used on the scheduler machine or is this a per hypervisor setting?

Thx,
Robert van Leeuwen