Dynamic Performance Forecasting for Network-Enabled Servers in a Heterogeneous Environment

Frédéric Desprez 1 Martin Quinson 1 Frédéric Suter 1
1 REMAP - Regularity and massive parallel computing
Inria Grenoble - Rhône-Alpes, LIP - Laboratoire de l'Informatique du Parallélisme
Abstract : This paper presents a tool for dynamic forecasting of Network-Enabled Servers performance. FAST (Fast Agent's System Timer}) is a software package allowing client applications to get an accurate forecast of communicat- ion and computation times and memory use in a heterogeneous environment. It relies on low level software packages, i.e., network and host monitoring tools, and some of our developments in computation routines modeling. The FAST internals and user interface are presented and a comparison between the execution time predicted by FAST and the measured time of complex matrix multiplication executed on an heterogeneous platform is given.
Document type :
Reports
Complete list of metadatas

https://hal.inria.fr/inria-00072267
Contributor : Rapport de Recherche Inria <>
Submitted on : Tuesday, May 23, 2006 - 8:15:54 PM
Last modification on : Friday, April 20, 2018 - 3:44:24 PM
Long-term archiving on : Sunday, April 4, 2010 - 11:01:14 PM

Identifiers

  • HAL Id : inria-00072267, version 1

Collections

Citation

Frédéric Desprez, Martin Quinson, Frédéric Suter. Dynamic Performance Forecasting for Network-Enabled Servers in a Heterogeneous Environment. [Research Report] RR-4320, INRIA. 2001. ⟨inria-00072267⟩

Share

Metrics

Record views

379

Files downloads

114