Analysis and Modeling of Social Influence in High Performance Computing Workloads

Shuai Zheng, Zon-yin Shae, Xiangliang Zhang, Hani Jamjoom and Liana Fong

17th International European Conference on Parallel and Distributed Computing (Euro-Par)

Bordeaux, France, August 2011

Abstract. Social inﬂuence among users (e.g., collaboration on a project) creates bursty behavior in the underlying high performance computing (HPC) workloads. Using representative HPC and cluster workload logs, this paper identiﬁes, analyzes, and quantiﬁes the level of social inﬂuence across HPC users. We show the existence of a social graph that is characterized by a pattern of dominant users and followers. This pattern also follows a power-law distribution, which is consistent with those observed in mainstream social networks. Given its potential impact on HPC workloads prediction and scheduling, we propose a fast-converging, computationally-efficient online learning algorithm for identifying social groups. Extensive evaluation shows that our online algorithm can (1) quickly identify the social relationships by using a small portion of incoming jobs and (2) can efficiently track group evolution over time.

Keywords. HPC, Social influence

Link. /publications/jamjoom-EUROPAR-11.pdf

Bibtex.

@inproceedings{jamjoom-EUROPAR-11,
author = {Shuai and Zheng and Zon-yin and Shae and Xiangliang and Zhang and Hani and Jamjoom and Liana and Fong},
title = {{Analysis and Modeling of Social Influence in High Performance Computing Workloads}},
booktitle = {17th International European Conference
                  on Parallel and Distributed Computing (Euro-Par)},
address = {Bordeaux, France},
month = {August},
year = {2011}
}