In order to reduce building energy consumption, we need a global, accurate view of the building. Many modern buildings have a sensing infrastructure that can be used to do fine-grained accounting and understand the complex interactions between systems and spaces. However, through our experience with an active, campus-wide Building Management System (BMS) system, we observe that BMSs are not well suited for this task. In buildings, there is a fundamental relationship between systems and spaces. The Integrated Sensor-Stream Storage System (IS4) captures this relationship by constructing a resource hierarchy that names measurement instruments relative to their context. This allows us to bring together systems and spaces through the instruments within them. Furthermore, IS4 keeps track of changes in the hierarchy over time. It is also designed to simplify data acquisition and exploits the naming structure for simpler, context-dependent queries. We have implemented IS4 and are currently using it to monitor Cory Hall at UC Berkeley as well as independent deployments at Samsung, Intel, and Lawrence Berkeley National Laboratory.