In cloud-computing environments, every service in an application is deployed as a different service instance. All service instances are deployed at a different level of end-to-end Quality of Service (QoS), which are described in the Service Level Agreement (SLA). To satisfy the given SLAs and end-to-end QoS requirements of an application, the application is required to optimize its deployment configuration of service instances. In this paper, a genetic algorithm is implemented, as proposed in literature, to solve this problem by searching for the optimal solution from a search space while satisfying the given SLA. The algorithm estimates the performance of an application by minimizing the latency allowing SLAs to be defined in a probabilistic manner. Simulation results demonstrate that the genetic algorithm implemented in this paper obtains the deployment configurations that satisfy the given SLA.