In computing, the trend toward moving from centralized to distributed approaches is happening alongside lots of discussion about scalability, load balancing, and performance issues. Meanwhile, for huge data centers to respond to the new requests of implemented clouds, scientific computing, and social media systems, new requirements are propounded. Centralized management, coordination, and control of resources in the data centers cannot respond to these requirements effectively, so devolved distributed control mechanisms have been devised to mitigate the issues, and their performance-oriented problems have been the subject of discussions and investigations.

With a network management theme, and using clear and illustrative statements, the load balancing problem for devolved controllers (LBDC), the matter of the paper, is explained. How to migrate the flow switches from probable saturating controller to other defined partitions to create or maintain essential equilibrium. Standard deviation of the load on the controllers is considered as the unbalanced comparator metric.

The problem is formulated, and its NP completeness is emphasized. It is converted to a solvable integer program, with some relaxation to the linear one. The feasible solution is discovered and its performance is evaluated. To provide an implementable solution, the centralized and distributed algorithms for switch migration are composed, discussed, refined, and compared. A performance evaluation of the proposed platforms indicates great achievements concerned with the configuration of different controller placements and functional policies.

With all of its complexity, the paper has a smooth and specific nature, with stepwise progress toward acquiring the results. Continuity of the fruitful discussion atmosphere, throughout the paper, is its admirable feature. The paper is a useful artifact, which could interweave the theoretical discussion with practical results masterfully.