Currently we load both quantitative (how much disk, RAM, etc) and qualitative (SSD vs. spinning, special chipsets, etc.) into the flavor that a user can request. As the number of qualitative features has been increasing, there is the phenomenon of “flavor explosion”, where each combination of compute resources with each possible qualitative feature results in an exponential growth in the number of required flavors. I have a spec [0] that is one of the first steps to address this by separating out qualitative features from flavors.
Since many deployments use flavor as the equivalent of a SKU, this means that billing will now have to be modified to account for flavor + features instead of flavor alone. I’d like to start a discussion to better understand the implications of these proposed changes so that we can come up with the best approach to balance the respective needs. [0] https://review.openstack.org/#/c/313784/ -- Ed Leafe _______________________________________________ OpenStack-operators mailing list [email protected] http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-operators
