It is widely acknowledged that modern information sytems require an ontological layer on top of data, associated with advanced reasoning mechanisms able to exploit the semantics encoded in ontologies. We focus here on ontology-based data access (OBDA), a new paradigm that seeks to take ontological knowledge into account when querying data. This paradigm is currently the subject of intense research in the database, knowledge representation and reasoning, and Semantic Web communities. Indeed, it is expected to have a major impact in many application domains, however some foundational issues need first to be adressed. In this context, we consider an emerging logical framework based on existential rules, also known as Datalog+/-. This framework can also be defined in graph terms. Compared to the lighweight description logics currently developed for OBDA, it is more powerful and flexible; an important feature is that predicate arity is not restricted, which allows for a natural coupling with database schemas and facilitates the integration of additional information, such as contextual knowledge. On the other hand, the existential rule framework extends the deductive database language Datalog by enabling to infer the existence of entities that do not necessarily occur in the database (hence the name existential rules), a feature that has been recognized as crucial in the context of incomplete information. In this talk, we will provide an introduction to this framework in the context of OBDA, then present the main decidability and complexity results as well as algorithmic techniques, and discuss some challenging research issues.