000165372 001__ 165372
000165372 005__ 20190316235103.0
000165372 037__ $$aREP_WORK
000165372 245__ $$aTuning Paxos for high-throughput with batching and pipelining
000165372 269__ $$a2011
000165372 260__ $$c2011
000165372 300__ $$a18
000165372 336__ $$aReports
000165372 520__ $$aPaxos is probably the most popular state machine replication protocol. Two optimizations that can greatly improve its performance are batching and pipelining. Nevertheless, tuning these two optimizations to achieve optimal performance can be challenging, as their effectiveness depends on many parameters like the network latency and bandwidth, the speed of the nodes, and the properties of the application. We address this question, by first presenting an analytical model of the performance of Paxos that can be used to obtain values for tuning batching and pipelining. We then present experiments validating the model and investigating how these two optimizations interact in both a LAN and a WAN setting. The results show that although batching by itself is usually sufficient to maximize the throughput in a LAN environment, in a WAN it must be complemented with pipelining.
000165372 6531_ $$aPaxos
000165372 6531_ $$aBatching
000165372 6531_ $$aPipelining
000165372 6531_ $$aAnalytical Study
000165372 6531_ $$aExperimental Evaluation
000165372 700__ $$aSantos, Nuno
000165372 700__ $$0241767$$aSchiper, André$$g106377
000165372 8564_ $$s350332$$uhttps://infoscience.epfl.ch/record/165372/files/santos11%20Tuning%20Paxos%20for%20high-throughput%20with%20batching%20and%20pipelining%20-%20TR_2.pdf$$zn/a
000165372 909C0 $$0252206$$pLSR$$xU10411
000165372 909CO $$ooai:infoscience.tind.io:165372$$pIC$$preport$$qGLOBAL_SET
000165372 917Z8 $$x175974
000165372 917Z8 $$x175974
000165372 937__ $$aEPFL-REPORT-165372
000165372 973__ $$aEPFL$$sPUBLISHED
000165372 980__ $$aREPORT