Optimising a storage network configuration using generic modification

摘要

A method is provided for determining low cost solutions to a storage area network problem. A plurality of possible network solutions are generated and the cost of each is calculated. In one preferred approach a "genetic" algorithm is applied such that lower cost solutions are bred together in preference to higher cost solutions and that these parent solutions are then evaluated for their cost during the next iteration. There is also described a method for optimising a network defining a plurality of flows through a set of fabric nodes, investigating the cost of establishing some routes through the network and updating a route desirability variable as a function of the cost.