Grids promote the publication, sharing and integration of scientific data, distributed across virtual organizations. The complexity of data management in a grid environment comes from the distribution, heterogeneity, dynamicity and number of data sources. Data virtualization is a fundamental issue to manage in a unified and virtualized manner (from structure, location, access service, performance points of view) data stored into multiple, geographically spread data sources. It represents a key point for distributed data management services which has to be addressed to build high quality production/enterprise oriented services. In this work we talk about the convergence process (among three main data grid services developed in the context of the GRelC Project) that has led to the unified GRelC data access and integration service.