Managing flash crowds on the Internet

Arı, İsmail; Hong, Bo; Miller, Ethan L.; Brandt, Scott; Long, Darrell D. E.

doi:10.1109/mascot.2003.1240667

Cited by 94 publications

(82 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Using an admission control mechanism to prevent additional overload has also been suggested [123]. On a more global scale, adequate caching mechanisms can alleviate the problem [33].…”

Section: Flash Crowdsmentioning

confidence: 99%

See 1 more Smart Citation

Workload Modeling for Computer Systems Performance Evaluation

Feitelson

2015

197

131

View full text Add to dashboard Cite

Reliable performance evaluations require the use of representative workloads. This is no easy task since modern computer systems and their workloads are complex, with many interrelated attributes and complicated structures. Experts often use sophisticated mathematics to analyze and describe workload models, making these models difficult for practitioners to grasp. This book aims to close this gap by emphasizing the intuition and the reasoning behind the definitions and derivations related to the workload models. It provides numerous examples from real production systems, with hundreds of graphs. Using this book, readers will be able to analyze collected workload data and clean it if necessary, derive statistical models that include skewed marginal distributions and correlations, and consider the need for generative models and feedback from the system. The descriptive statistics techniques covered are also useful for other domains.

show abstract

“…Using an admission control mechanism to prevent additional overload has also been suggested [123]. On a more global scale, adequate caching mechanisms can alleviate the problem [33].…”

Section: Flash Crowdsmentioning

confidence: 99%

“…Slightly lower down we start seeing navigational queries targeting more traditional companies and institutions, such as "walmart" (22), "southwest airlines" (31), and "target" (33). These typically have only one main variant, without Internet address trimmings.…”

Section: Transactional Queriesmentioning

confidence: 99%

Workload Modeling for Computer Systems Performance Evaluation

Feitelson

2015

197

131

View full text Add to dashboard Cite

show abstract

“…The flashcrowd starts at the beginning of the increase and finishes at the end of the increase. Similar definitions are also used [3], [7] in modeling web server workloads. As illustrated in Figure 1, our flashcrowd model consists of four components: the arrival time, which is the time between the creation of a swarm and the start of a flashcrowd; the duration, which is the time between the start and the end of the increase; the plateau period, which is the time period with limited churn immediately after a flashcrowd ends; and the magnitude, which indicates the significance of a flashcrowd in terms of the increase.…”

Section: A Model Of Bittorrent Flashcrowdsmentioning

confidence: 99%

Identifying, analyzing, and modeling flashcrowds in BitTorrent

Zhang

Iosup

Pouwelse

et al. 2011

2011 IEEE International Conference on Peer-to-Peer Computing

View full text Add to dashboard Cite

Abstract-Flashcrowds-sudden surges of user arrivals-do occur in BitTorrent, and they can lead to severe service deprivation. However, very little is known about their occurrence patterns and their characteristics in real-world deployments, and many basic questions about BitTorrent flashcrowds, such as How often do they occur? and How long do they last?, remain unanswered. In this paper, we address these questions by studying three datasets that cover millions of swarms from two of the largest BitTorrent trackers. We first propose a model for BitTorrent flashcrowds and a procedure for identifying, analyzing, and modeling BitTorrent flashcrowds. Then we evaluate quantitatively the impact of flashcrowds on BitTorrent users, and we develop an algorithm that identifies BitTorrent flashcrowds. Finally, we study statistically the properties of BitTorrent flashcrowds identified from our datasets, such as their arrival time, duration, and magnitude, and we investigate the relationship between flashcrowds and swarm growth, and the arrival rate of flashcrowds in BitTorrent trackers. In particular, we find that BitTorrent flashcrowds only occur in very small fractions (0.3-2%) of the swarms but that they can affect over ten million users.

show abstract

“…Consequently, the longest jobs in this trace are limited to 18 hours (since a job that exceeds its runtime estimate is killed by the system). However, in a real system, it takes some time to propagate the instruction to kill a job to all the nodes 2 . Therefore the trace indicates that some jobs run for a bit more than 18 hours.…”

Section: Example Of a Butterfly Effectmentioning

confidence: 99%

“…Of course, in these particular cases, high-load conditions may be more important and meaningful than normal conditions; if this is the case, they should be the focus of study rather than being eliminated as suggested below. For example, Ari et al model such activity, which they call "flash crowds", with the aim of evaluating schemes to survive them [2].…”

Section: Generalizingmentioning

confidence: 99%