With the ubiquity of large-scale graph data in a variety of application domains, querying them effectively is a challenge. In particular, reachability queries are becoming increasingly important, especially for containment, subsumption, and connectivity checks. Whereas many methods have been proposed for static graph reachability, many real-world graphs are constantly evolving, which calls for dynamic indexing. In this paper, we present a fully dynamic reachability index over dynamic graphs. Our method, called DAGGER, is a light-weight index based on interval labeling, that scales to million node graphs and beyond. Our extensive experimental evaluation on real-world and synthetic graphs confirms its effectiveness over baseline methods.