달력

92010  이전 다음

  •  
  •  
  •  
  • 1
  • 2
  • 3
  • 4
  • 5
  • 6
  • 7
  • 8
  • 9
  • 10
  • 11
  • 12
  • 13
  • 14
  • 15
  • 16
  • 17
  • 18
  • 19
  • 20
  • 21
  • 22
  • 23
  • 24
  • 25
  • 26
  • 27
  • 28
  • 29
  • 30
  •  
  •  

Ganesha: Black-Box Diagnosis of MapReduce Systems
Link:http://portal.acm.org/citation.cfm?id=1710115.1710118
Xinghao Pan, Jiaqi Tan, Soila Kavulya, Rajeev Gandhi, Priya Narasimhan   Carnegie Mellon University, Pittsburgh, PA
SPECIAL ISSUE: Special issue on the workshop on hot topics in measurement & modeling of computer systems (HotMetrics 2009)

Abstract -
Ganesha aims to diagnose faults transparently (in a black-box manner) in MapReduce systems, by analyzing OS-level metrics. Ganesha's approach is based on peer-symmetry under fault-free conditions, and can diagnose faults that manifest asymmetrically at nodes within a MapReduce system. We evaluate Ganesha by diagnosing Hadoop problems for the Gridmix Hadoop benchmark on 10-node and 50-node MapReduce clusters on Amazon's EC2. We also candidly highlight faults that escape Ganesha's diagnosis.

[문제 설명]
There are manu challenges in problem localization and root-cause analysis.
[목표를 설명할 때]
Ganesha aims to diagnose faults transparently in MapReduce systems, by analyzing OS-level metrics.
[설명]
We describe Ganesha, our black-box diagnostic approach that we apply to diagnose such performance problems in Hadoop.
[메뉴 배열]
3.1 Hypotheses, 3.2 Goals, 3.3 Non-Goals, 3.4 Assumptions
[수식 설명]
Thus, the true-positive and false-positive ratios are computed as : [수식]

Posted by Teshi
Paper Link : ACM Portal

Thomas Karagiannis, Konstantina Papagiannaki, Michalis Faloutsos
ACM SIGCOMM Computer Communication Review,
Volume 35 ,  Issue 4  (October 2005), SESSION: Security, Pages: 229 - 240,Year of Publication: 2005, ISSN:0146-4833

 In this paper, they thought traffic classifiers of the future will need to classify traffic "in the dark" by application trends and the increasing use of encryption. So they suggest different approach to classifying traffic flow according to the applications that generate them. In contrast to pervious methods, there approach is based on observing and identifying patterns of host behavior at the transport layer. They suggest analyze these patterns at three levels, social, functional and application level.

 1. Taking into account empirical application trends ant the increasing use of encryption, we conjecture that traffic classifiers of the future will need to classify traffic "in the dark".
2. Furthermore, our approach has two important features.
3. In contrast to previous methods, our approach is based on observing and identifying patterns of host behavior at the transport layer.

Posted by Teshi