The considered problem comes down to searching for leasing offers that balance the needs of customers with the capabilities of the PaaS provider. In the considered problem instance a particular role is played by the costs of loss of the offered functionalities of leased equipment caused by its failures. Therefore, the planned offers should take into account the associated risks. The proposed concept of the so-called the offers robustness specifying the degree of reduction in costs incurred by the customer as a result of the failure of a specific device will make this possible. In turn, a greater offer robustness, implying redundancy in the functionality of the rented devices, causes its price to increase. Therefore, the requested offer should minimize both the costs of rented equipment and the risk of its failure. The proposed approach using a genetic algorithm is illustrated with practical examples.