Public cloud service providers provide an infrastructure that gives businesses and individuals access to computing power and storage space on a pay-as-you-go basis. This allows these entities to bypass the usual costs associated with having their own data centre such as: hardware, construction, air conditioning and security costs, for example, making this a cost-effective solution for data storage. If the data being stored is of a sensitive nature, encrypting it prior to outsourcing it to a public cloud is a good method of ensuring the confidentiality of the data. With the data being encrypted, however, searching over it becomes unfeasible. In this paper, we examine different architectures for supporting search over encrypted data and discuss some of the challenges that need to be overcome if these techniques are to be engineered into practical systems.