Efficient fault monitoring with Collaborative Prediction

Dawei Feng 1 Cecile Germain-Renaud 1, 2 Tristan Glatard 3
2 TAO - Machine Learning and Optimisation
INRIA Saclay - Ile de France, UP11 - Université Paris-Sud - Paris 11, CNRS : UMR8623, LRI - Laboratoire de Recherche en Informatique
Abstract : Isolating users from the inevitable faults in large distributed systems is critical to Quality of Experience. We formulate the problem of probe selection for fault prediction based on end-to-end probing as a Collaborative Prediction (CP) problem. On an extensive experimental dataset from the EGI grid, the combination of the Maximum Margin Matrix Factorization approach to CP and Active Learning shows excellent performance, reducing the number of probes typically by 80% to 90%.
Document type :
Conference papers
Journées scientifiques mésocentres et France Grilles, Oct 2012, Paris, France. 2012, <http://mesogrilles2012.sciencesconf.org/8516/document>


https://hal.inria.fr/hal-00758025
Contributor : Cecile Germain-Renaud <>
Submitted on : Tuesday, November 27, 2012 - 9:30:56 PM
Last modification on : Monday, February 16, 2015 - 1:03:45 PM

File

mesogrilles_CPFDV2.pdf
fileSource_public_author

Identifiers

  • HAL Id : hal-00758025, version 1

Collections

Citation

Dawei Feng, Cecile Germain-Renaud, Tristan Glatard. Efficient fault monitoring with Collaborative Prediction. Journées scientifiques mésocentres et France Grilles, Oct 2012, Paris, France. 2012, <http://mesogrilles2012.sciencesconf.org/8516/document>. <hal-00758025>

Export

Share

Metrics

Consultation de
la notice

177

Téléchargement du document

58