Research Article
Evolutionary Multiobjective Query Workload Optimization of Cloud Data Warehouses
Algorithm 2
Multiobjective optimization of cloud database configuration using genetic algorithm.
| Input: Set of VM types (VMs), Query workload (W) | | Output: A set of pareto-optimal solutions | | () VM: Set of Virtual Machine types | | () QP: Set of alternative query plans | | () N: Set of alternative network bandwidths | | () p: Population | | () , : Individuals (parent) selected for crossover or mutation | | () s: Generated individual | | () p Generate random individuals(VM, QP, N) | | () Calculate fitness of individuals(p) | | () p truncate(p) | | () B_response_time Find_best_response_time(VMs, W) | | () Cheapest_cost Find_cheapest_cost(VMs, W) | | () for to generations do | | () (, ) Select pair of parents(p) | | () s Crossover(, ) | | () Replace with least-fit in the population(p, s) | | () s Mutation(p, s) | | () Replace with least-fit in the population(p, s) | | () Replace duplicate_chromosomes(p) | | () Update B_response_time | | () Update Cheapest_cost |
|
