On Tue, Apr 14, 2009 at 1:12 PM, Mehdi Sheikhalishahi < mehdi.alishahi@gmail.com> wrote:
Dear All:
In the API Specification, Do you address the scheduling parameters for Advance Reservation, Co-allocation and others?
Interesting question... I have spent a large chunk of my weekend roughing up the API<http://forge.ogf.org/sf/wiki/do/viewPage/projects.occi-wg/wiki/APIDesign>and rather than bundling everything in together have so far opted for a core + extensions approach. This fits nicely with wanting to support everything from single-user virtualised desktops and hypervisors burnt into motherboards to global infrastructures with millions of resources and a single entry point. For clouds (like Amazon EC2) that have fixed hardware configurations rather than fluid resource allocations I'm planning to expose templates. One way to implement Amazon EC2 style reservations would be to implement a "reserve" actuator similar to "start", "stop" and "restart". This also touches on the question of whether instances persist when they are stopped - ephemeral implementations should be supported as well so this is something I'm thinking about. As Andrew said, best thing for you to do is explain your requirements in a use case. Sam