System and method for a self-optimizing reservation in time of compute resources

A method comprising:

• performing the following in a compute environment comprising a first group of compute resources and a second group of compute resources:
• receiving from a requestor, at a first time, a request for a reservation of compute resources in the compute environment, the request comprising a preferred criteria and a required criteria;
• establishing the reservation of the first group of compute resources to meet the required criteria such that a second time is set for a workload to use the first group of compute resources, wherein the second time is after the first time;
• prior to the second time, determining if the preferred criteria can be met by migrating the reservation from the first group of compute resources to the second group of compute resources;
• in response to determining that the preferred criteria can be met by migrating the reservation from the first group of compute resources to the second group of compute resources, migrating the reservation from the first group of compute resources to the second group of compute resources;
• and
• in response to determining that the preferred criteria cannot be met by migrating the reservation from the first group of compute resources to the second group of compute resources:
• determining, after the second time, that the preferred criteria can be met by migrating the reservation from the first group of compute resources to the second group of compute resources;
• comparing a cost of cancelling the reservation of the first group of compute resources and reserving the second group of compute resources with a threshold, wherein the threshold indicates when overall resource efficiency can be improved in that enough of the preferred criteria can be met to overcome the cost;
• and
• in response to the cost of cancelling the reservation being less than the threshold, migrating the reservation from the first group of compute resources to the second group of compute resources by canceling the reservation for the first group of resource and reserving the second group of compute resources for the workload to use the second group of compute resources to meet the preferred criteria;
• and
• providing an access list on the reservation that governs access to the second group of compute resources.