Proceedings. 20th International Conference on Data Engineering (2004)
Mar. 30, 2004 to Apr. 2, 2004
Roger Barga , Microsoft Research, Redmond, WA
Shimin Chen , Carnegie Mellon University, Pittsburgh, PA
David Lomet , Microsoft Research, Redmond, WA
Phoenix/App supports software components whose states are made persistent across a system crash via redo recovery, replaying logged interactions. Our initial prototype force logged all request/reply events resulting from inter-component method calls and returns. This paper describes an enhanced prototype that implements: (i) log optimizations to improve normal execution performance; and (ii) checkpointing to improve recovery performance. Logging is reduced in two ways: (1) we only log information required to remove non-determinism, and we only force the log when an event "commits" the state of the component to other parts of the system; (2) we introduce new component types that provide our enhanced system with more information, enabling further reduction in logging. To improve recovery performance, we save the values of the fields of a component to the log in an application "checkpoint". We describe the system elements that we exploit for these optimizations, and characterize the performance gains that result.
S. Chen, D. Lomet and R. Barga, "Improving Logging and Recovery Performance in Phoenix/App," Proceedings. 20th International Conference on Data Engineering(ICDE), Boston, Massachusetts, 2004, pp. 486.