Abstract: A method in a server end station is described. The method includes performing an iteration of a rebalancing computation by selecting a set of one or more service sets for rebalancing, wherein the selecting the set of one or more service sets is based on service set constraints and host constraints; generating candidate solutions, wherein each candidate solution includes a randomized one-to-one mapping of each of the service sets to one of the hosts; performing one or more crossover operations on the candidate solutions; performing one or more mutation operations on the additional candidate solutions; selecting as a solution one of the candidate solutions that has a best fitness score, wherein a fitness score for a candidate solution is calculated based on the distribution of resources resulting from and number of migrations needed for the candidate solution; and repeating the iteration of the rebalancing computation an additional number of times.

Abstract: A method in a server end station is described. The method includes performing an iteration of a rebalancing computation by selecting a set of one or more service sets for rebalancing, wherein the selecting the set of one or more service sets is based on service set constraints and host constraints; generating candidate solutions, wherein each candidate solution includes a randomized one-to-one mapping of each of the service sets to one of the hosts; performing one or more crossover operations on the candidate solutions; performing one or more mutation operations on the additional candidate solutions; selecting as a solution one of the candidate solutions that has a best fitness score, wherein a fitness score for a candidate solution is calculated based on the distribution of resources resulting from and number of migrations needed for the candidate solution; and repeating the iteration of the rebalancing computation an additional number of times.