The Case for Evaluating MapReduce Performance Using Workload Suites

By Yanpei Chen, Archana Ganapathi, Rean Griffith, Randy Howard Katz. 2011 IEEE 19th Annual International Symposium on Modelling, Analysis, and Simulation of Computer and Telecommunication Systems, 2011

458 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Mirror mirror on the ceiling: flexible wireless links for data centers (2012, background): “Most prior work addresses traffic congestion through network architecture design and traffic scheduling [11, 13, 14, 18, 21, 22], or modeling network traffic characteristics [15, 16, 25, 17].”
Toward Scalable Systems for Big Data Analytics: A Technology Tutorial (2014, background): “[288] developed an open source statistical workload injector for MapReduce (SWIM).”
Interactive Analytical Processing in Big Data Systems: A Cross-Industry Study of MapReduce Workloads (2012, methodology): “There are some inaccuracies at trace start and termination, due to partial information for jobs straddling the trace boundaries.”
Decentralized task-aware scheduling for data center networks (2014, methodology): “We use network over-subscription of 2:1 at the ToR switch and a selectivity of 3% (data input-to-output ratio for aggregator nodes), which is consistent with observations of live systems for data aggregation tasks [9].”
Parasol and GreenSwitch: managing datacenters powered by renewable energy (2013, background): “We study two widely different Hadoop traces, called Facebook and Nutch .”