Christophe Cérin Home Page

— The effective autonomy is the confidence in researchers and their freedom of initiative —
— Evaluation does not mean 'Credit Rating Agency' —

Jan 17, 2011

Introduction

— Posted by Christophe Cérin @ 9:57 am —

This page aims to provide information about my professional activities (mainly) as full time professor at the university of Paris XIII. My research is with Laboratoire d'Informatique de Paris Nord [LIPN]. Member of a team-project with the sino-french LIAMA laboratory, located in Beijing.

I'am also involved with teaching activities at IUT de Villetaneuse, département d'informatique. My principal teaching topic is Operating Systems (introductory course -- in french). I'am chairing the Licence PRO SIL for the department of vocational studies.

My own teaching activity is centered on basic knowledge for fresh students. I'am also concerned about the relationship between universities and companies. For all of that, I maintain the Stubok. Feel free to send me back any comment on this book in French.

In the past, I was involved with the following courses: Innovative Technology, PostScript, BSP, Python, SVG [see here] and Introduction to computer architecture [see here].

August 19, 2011

Research Interests

— Posted by Christophe Cérin @ 10:53 am —

The "Distributed Algorithms, Software and Architecture" research group is working on:

  • Software architecture of large scale systems; Publication / Subscription paradigm;
  • Models of large scale systems; computation of fault tolerant properties on such systems;
  • Middleware for testing algorithms inside of a generic execution support for large scale systems;
  • Numerical parallel algorithms;
  • Parallel programming language for shared memory architecture;

My work is about Parallel Computing in the field of middleware for Grids and Clouds and also in the field of High Performance Computing (HPC).

In the field of Grids, we build fully distributed middleware based on the Desktop Grid paradigm and the challenges are to coordinate the different entities participating into the computation or into the billing. A paper for HPDC'2011 is also comparing our work with other projects. Two examples of our results are:

BonjourGrid is an example of a middleware to coordinate multiple instances of desktop grids that we are developing in my research group in the context of an international collaboration with Tunisia. A BonjourGrid virtual machine based on Xen for the use with Boinc, Condor and XtremWeb can be downloaded from Sourceforge; please, send me an email is you need more information.

Event-based.org portal -- everything events! BonjourGrid is implemented according to the pub-sub paradigm including the fault tolerant part that manage the crash of coordinators. Pub-sub is an example of event-based programming. BonjourGrid is also based on Bonjour from Apple, a Local Area Network protocol running on your Mac. BonjourGrid ideas are the core of an industrial partnership starting on September 2011. The key idea of our work is to adapt the concepts of BonjourGrid for a distributed release of the SlapOS orchestrator. The project has been accepted under the framework of "Pôles de compétitivité AAP-R et D-FUI n°12". The project is, in part described on the Nexedi web site.

PastryGrid is another desktop Grid middleware able to execute 'on the fly' task graphs with dependencies as well as bag of tasks. PastryGrid is built on top of Pastry, it scales well and is also fault-tolerant.

Currently, I'am also working in the field of HPC for building execution supports that scale. The general problem is that since the number of cores in a parallel machine will reach more than 1 million in a near future, faults can be handled as soon as possible in the design of software running on top of parallel machines. Recently (SBAC'2011), we proposed to modelize a logical networks for controlling any large scale platforms according to the notion of graph (Binomial Graphs and variant in our experiments) and we designed efficient algorithms for computing important metrics (average distance, message density, node connectivity, link connectivity, diameter and fault diameter). Typically, computing all these metrics lasts less than 5 minutes for a large graph of 21000 nodes. For obtaining such performance, we use probabilistic algorithms that approximate the result in a good way.

I'am pleased to invit you to submit a paper for a major IEEE federation of conferences related to SERVICES. Visit http://www.servicescongress.org/2012/
2012 IEEE World Congress on Services will serve as a federation to host the FIVE co-located theme topic conferences to explore the deep knowledge space of Services Computing in different directions:
  • http://www.thecloudcomputing.org/2012/-IEEE 5th International Conference on Cloud Computing (CLOUD 2012) will focus on modeling, developing, publishing, monitoring, managing, delivering XaaS in the context of various types of cloud environments. (http://www.thecloudcomputing.org/2012/)
  • http://conferences.computer.org/icws/2012/- IEEE 19th International Conference on Web Services (ICWS 2012) will focus on data-centric Web-based services, featuring Web services modeling, development, publishing, discovery, composition, testing, adaptation, and delivery, service-oriented architecture (SOA), and Web services technologies as well as standards. (http://conferences.computer.org/icws/2012/)
  • http://conferences.computer.org/scc/2012/- IEEE 9th International Conference on Services Computing (SCC 2012) will focus on services innovation lifecycle that includes enterprise modeling, business consulting, solution creation, services orchestration, services optimization, services management, services marketing, business process integration and management. (http://conferences.computer.org/scc/2012/)
  • http://www.themobileservices.org/2012/-IEEE 1st International Conference on Mobile Services (MS 2012) will focus on the development, publication, discovery, orchestration, invocation, testing, delivery, and certification of mobile applications and services. (http://www.themobileservices.org/2012/)
  • http://www.ieeese.org/2012/-IEEE 1st International Conference on Services Economics (SE 2012) will focus on economic and enterprise transformation aspects of the utility-oriented services paradigm. (http://www.ieeese.org/2012/)

Facts:

May 3rd, 2009

Administration of Research

— Posted by Christophe Cérin @ 10:35 am —

Local chair for the Resilience project (Pôle de compétitivité -- 2011- 2013).

Chair of SafeScale Project (Security And Fault-tolerance to Exploit Safety ambient Computing in lArge scaLe Environments - 2006- 2008) funded by ANR (Agence Nationale de la Recherche). Organizer of SGS workshop with Europar 2008 (see the LNCS proceedings with Springer here). General Chair of GPC 2007 (Grid and Pervasive Computing) conference (see the LNCS proceedings with Springer here).

Member of the Steering Committee of GPC conference series. Member of the program committees of CCGRID, SBAC, AINA conferences. Cooperation with China (Wuhan, Beijing, Hangzhou).

September 3rd, 2008

Leisure, Sport...

— Posted by Christophe Cérin @ 6:54 pm —

If you are asking for nice cycling tours between "Cap Nègre" and "Fort de Brégançon" (Var - France), please contact me. I can make special arrangements for you (thanks to Michel) in order to pass along "Col du Babaou", close to "Collobrières" village. The tours are free of harmful politicians and fashion victims. I'am sure that you will meet only respectable people!

I have also an experience of cycling in "Bois de Vincennes" (Paris) until Créteil and "bords de Marne". Don't hesitate to contact me on this matter.

Release number: 3.14116 || Inspired from a css template by WordPress