Cover Fe Ature

37 FEBRUARY 2012 Published by the IEEE Computer Society 0018-9162/12/$31.00 © 2012 IEEE capabilities and in exposing the exaggerated marketing hype of many commercial DDBSs. However, since its initial formal proof,7 CAP has become increasingly misunderstood and misapplied, potentially causing significant harm. In particular, many designers incorrectly conclude that the theorem imposes certain restrictions on a DDBS during normal system operation, and therefore implement an unnecessarily limited system. In reality, CAP only posits limitations in the face of certain types of failures, and does not constrain any system capabilities during normal operation. Nonetheless, the fundamental tradeoffs that inhibit DDBSs’ capabilities during normal operation have influenced the different design choices of well-known systems. In fact, one particular tradeoff—between consistency and latency—arguably has been more influential on DDBS design than the CAP tradeoffs. Both sets of tradeoffs are important; unifying CAP and the consistency/latency tradeoff into a single formulation—PACELC—can accordingly lead to a deeper understanding of modern DDBS design.