Starfish: A Self-tuning System for Big Data Analytics

By Herodotos Herodotou, Harold Lim, Gang Luo, N. Borisov, Liang Dong, F. Cetin, Shivnath Babu. Conference on Innovative Data Systems Research, 2011

722 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/660fa773b70629aa8c36f4e11812b2b07fcfa04d

How researchers see this paper

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

Citations in the corpus

Profiling, what-if analysis, and cost-based optimization of MapReduce programs (2011, background): “Personal communication, our own experience [3, 15], and plenty of anecdotal evidence on the Web indicate that finding good configuration settings for MapReduce jobs is time consuming and requires extensive knowledge of system internals.”
The only constant is change: incorporating time-varying network reservations in data centers (2012, methodology): “PROTEUS shares the same pro.ling methodology with Elastisizer [22], StarFish [23] and CBO [21] which focus on choosing the type and number of VMs for MapReduce jobs to balance cost/performance objectives.”
SkewTune: mitigating skew in mapreduce applications (2012, methodology): “proposed the Starfish optimization framework that hierarchically optimizes from MapReduce jobs to workflows by searching for good parameter configurations [15].”
Big Data computing and clouds: Trends and future directions (2013, background): “Starfish [72], a data analytics system built atop Hadoop, focuses”
Data-centric Artificial Intelligence: A Survey (2023, unknown): “For instance, Starfish [96] proposes a profile-predict-optimize approach that generates job profiles with dataflow and cost statistics, which are then used to predict virtual job profiles for task scheduling.”
Analytics over large-scale multidimensional data: the big data revolution! (2011, background): “Basically, this innovative paradigm aims at integrating the classical, well-known benefits of multidimensional data models (such as multidimensional abstractions, hierarchy-based dimensional tables, multi-resolution fact tables, multi-way aggregations, OLAP tools, and so forth) with analytics, in or”
Ernest: Efficient Performance Prediction for Large-Scale Advanced Analytics (2016, methodology): “Projects like MRTuner [68] and Starfish [48] model MapReduce jobs at very fine granularity and set optimal values for options like memory buffer sizes etc.”
Big Data Analytics: A Literature Review Paper (2014, unknown): “m, thus making it popular for big data analytics by load uted file system, and running parallel MapReduce com gets its magnetism and agility from the fact that dat by copying files into the distributed file system, and M at processing time rather than loading time [11].”