Abstract. A new algorithm is presented for detecting whether a particular computation of an asynchronous distributed system satisfies Poss Φ (read “possibly Φ”), meaning the...
Existing approaches to debugging distributed systems involve a cycle of passive observation followed by computation replaying. We propose predicate control as an active approach t...
Distributed programs are hard to write. A distributed debugger equipped with the mechanism to re-execute the traced computation in a controlled fashion can greatly facilitate the ...
Interconnect delays are increasingly becoming the dominant source of performance degradation in the nano-meter regime, largely because the wires do not scale as fast as the device...
Ray Robert Rydberg III, Jabulani Nyathi, Jos&eacut...