"The implication is that data is too extensive to be queried and moved into another resource for analysis, so the analysis program is instead distributed to the data-holding resources, with only the results being aggregated on a remote resource. This concept of data locality is actually a critical aspect of parallel data architectures. Additional system concepts are the interoperability (ability for tools to work together), reusability (ability to apply tools from one domain to another), and extendibility (ability to add or modify existing tools for new domains). These system concepts are not specific to Big Data, but their presence in Big Data can be understood in the examination of a Big Data reference architecture. . . ."