The pricing and capacity planning problem of IaaS cloud products has received widespread attention in recent years, however, this problem has not been studied from the perspectives of complementary and multiplexed resources as well as bundle sales. Considering an IaaS provider with two complementary resources and two cloud products (reserved instances and spot instances), and incorporating resource multiplexing into resource capacity planning, this paper first compares optimal pricing, capacity planning decisions and profits of the IaaS provider in three sales models of reserved instances (individual sales, bundle sales, and hybird sales), and then investigates both the IaaS provider and the cloud users" preferred sales models. The results show that: (1) For the IaaS provider, its profit reaches maximum when it selects the model of hybird sales, if only the models of individual sales and bundle sales are allowed, then when both the degree of complementary and the bundling discount price sensitivity are relatively large, it would select the model of bundle sales, otherwise, it would select the model of individual sales. (2) For cloud users, bundled product users prefer the model of bundle sales, and unbundled product users prefer the model of hybird sales. (3) The smaller the degree of complementary of the products, the greater the bundle discount. (4) Multiplexing resources makes the price of reserved instances lower, and the demand of reserved instances and the profit of the IaaS provider greater, therefore, it enables a win-win situation for both the IaaS provider and cloud users. This paper provides theoretical support and practical suggestions for bundle pricing and capacity planning decisions of IaaS cloud products with complementary and multiplexed resources.