Mesaac Makpangou

I opened my eyes in 1960, in Foumban (Cameroon). I am married and am a father of three daughters.

I received a Licence in Mathematics from University of Yaoundé (Cameroon), in june 1981. I did some teaching in 1982 at high schools at Yaoundé. In january 1983, thanks to a scolarship from the cameroonian government, I moved to France and then switched from Mathematics to Computer Science. I received a Ph.D in Computer Science from Paris 6 University (France), in february 1989. I did a post-doc at Cornell, in the ISIS group with Professor Kenneth Birman, from may 1989 to march 1990.

Since 1990, I hold a position of Researcher for the French National Institute for Research in Computer Science and Control (INRIA). I am currently member of Regal (Répartition et Gestion des Applications à Large Echelle) group, a common project of INRIA Rocquencourt and Paris 6 University. I worked previously within the INRIA SOR Group. I led the SOR group from 1999 through its replacement by the Regal INRIA and Paris 6 project in 2002.

Education

HDR (Habilitation à Diriger des Recehrches) in Computer Science from Paris 6 University (France), october 2001
Ph.D in Computer Science from Paris 6 University (France), february 1989
DEA (Diplôme d'Études Approfondies) in Distributed Systems from Paris 6 University (France), june 1985
Maîtrise in Computer Science from Paris 6 University (France), june 1984
Licence (akin Bachelor) in Computer Science from Paris 6 University (France), september 1983
Licence (akin Bachelor) in Mathematics from the University of Yaoundé (Cameroon), june 1981

Research Interest

I am interested in distributed systems. My two current projects are:

Deal an Adaptive Replication Middleware for Application Delivery Networks.
Decodes a Content Distributed Network for Web Applications.

Deal project

The Deal project aims to build an Adaptative Application Delivery Network (A-ADN) capable to adapt the performance target and the consistency guarantees for each hosted application, such as to match the specifics of applications and the objectives their providers.
We focus our work primarely on the design of an adaptive replication middleware for application delivery networks. The Deal adaptive replication middleware comprises four building blocks:

means to help applications providers specify the consistency constraints that underly the correction of services delivered by their applications and to indicate the performance target to pursue by the Delivery Network.
a generic and adaptable consistency management system that can be configured to match the specifics and the consistency constraints of any application.
an adaptive replication policy capable to take into account the specifics of each application, the performance target to pursue, the dynamic of the supporting infrastructure, and the requets distribution.
a distributed metrology system that collect or compute pertinent metrics at replica host servers, at clients sites, and on the heart of the network. Collected and computed metrics are then used to guide estimations and decisions to be made by the adaptive replication policy.

Decodes project

The Decodes (in French - Distribution Efficace des Contenus Dynamiques et Statiques) project aims to implement a CDN for web applications. This project extends our previous project in web caching and web monitoring. Within this project, we tackle three issues: the automatic fragmentation of web applications, fragment caching, and web application replication.

Previous projects

I have been previously involved in several other projects. These include:

Relais a directory system for CDNs.
Pandora a general purpose monitoring system that provides support for the construction of autonomic applications.
Boar a library of adaptable replicated objects.
SOS The Somiw Object-Oriented Operating System.

Publications

Here are my publications grouped by topics:

Adaptive Replication Middleware and replica divergence control protocol
Web caching and content distribution network
Web and Internet monitoring
Distributed shared objects
Some stuff for wide area distributed applications

Adaptive Replication Middleware

A Response Time Driven Replica Server Selection Substrate for Application Replica Hosting Systems. Corina Ferdean and Mesaac Makpangou. International Symposium on Applications and the Internet (SainT 2006), Phoenix, Arizona (USA), January 2006.
Efficient and Transparent Web-Services Selection. Nicolas Gibelin and Mesaac Makpangou. International Conference on service oriented computing (ICSOC 05).
A Fine-grained Customizable Consistency Protocol for Replicated Data Objects. Corina Ferdean and Mesaac Makpangou. Journées Francophones sur la Cohérence des Données en Univers Réparti (CDUR 2005), Paris (France), novembre 2005.
Exploiting Application Workload Characteristics to Accurately Estimate Replica Server Response Time. Corina Ferdean and Mesaac Makpangou. On the Move to Meaningful Internet Systems: DOA 2005, Agia Napa, Cyrus, october 2005.
A Generic and Flexible Model for Replica Consistency Management. Corina Ferdean and Mesaac Makpangou. International Conference on Distributed Computing and Internet Technology (ICDCIT 2004), Bhubaneswar, India, december 2004.
The Contact Substrate for Adaptive Replication of Internet Services. Corina Ferdean, Mesaac Makpangou, and Nicolas Gibelin. International Conference on Intelligent Access to the Multimedia Documents on the Internet (Medianet 2004), Tozeur, Tunisia, november 2004.
Flexible Contract Substrate for an Adaptable Replication System. Corina Ferdean and Mesaac Makpangou. Concurrent Information Processing and Computing Advanced RFesearch Workshop, Simaia, Roumania, july 2003.
A Scalable Replica Selection Strategy Based on Flexible Contracts. Corina Ferdean and Mesaac Makpangou. Third IEEE Workshop on Internet Applications (WIAPP03), San José, California (USA), june 2003.
Replica Divergence Control Protocol Based on Predicted Profile. Ahmed Jebali and Mesaac Makpangou. The International Conference on Parallel and Distributed Processing Techniques and Applications. Las Vegas, Nevada (USA), june 2002.
Replica Divergence Control Protocol in Weakly Connected Environment. Ahmed and Jebali Mesaac Makpangou. The IEEE International Symposium on Network Computing and Applications, Cambridge, MA (USA), october 2001.

Web Content Distribution

Caching Dynamic Content with Automatic Fragmentation. Ikram Chabbouh and Mesaac Makpangou. Seventh International Conference on Onformation Integration and Web-Based Applications and Services (IIWOS 05), Kuala Lumpur, Malaysia, september 2005.
Semantic-Aware Replication Protocol to Efficiently Share Large Web Community Document Store. Ahmed Jebali and Mesaac Makpangou. International Conference on Intelligent Access to the Multimedia Documents on the Internet (Medianet 2004), Tozeur, Tunisia, november 2004.
Capture and Exploitation of Dependencies and Events for Dynamic Web Content Invalidation. Ikram Chabbouh and Mesaac Makpangou. International Conference on Intelligent Access to the Multimedia Documents on the Internet (Medianet 2004), Tozeur, Tunisia, november 2004.
A Configuration Tool for Caching Dynamic Pages. Ikram Chabbouh and Mesaac Makpangou. International Workshop on Web Caching and Content Distribution, Beijing China, october 2004.
Syfax: un système de cache Web sémantique pour des communautés distribuées. Habib Smei, Mesaac Makpangou, and Abdelmajid Ben Hamadou. Journées francophones d'accès intelligent aux documents multimedia sur Internet, Sousse (Tunisie), juin 2002.
Reliable Directory Service for Weakly Consistent Replicated Caches. Mesaac Makpangou, Guillaume Pierre, Christian Khoury, and Neilze Dorta. International Conference on Distributed Computing Systems (ICDCS'99), june 1999.
Relais: un protocole de maintien de cohérence pour des caches web coopérants. Mesaac Makpangou et Éric Bérenguier. Conférence sur les Nouvelles Technologies pour la Répartition (NoTere'97), Pau (France), novembre 1997.

Web and internet monitoring

Pandora: an Efficient Platform for the Construction of Autonomic Applications. Simon Patarin et Mesaac Makpangou. In Self-Star Properties in Complex Information Systems, Lecture Notes in Computer Science, Hot Topics, Volume 3460, may 2005.
Pandora: une plate-forme efficace pour la construction d'applications autonomes. Simon Patarin et Mesaac Makpangou. Conférence francophone sur le déploiement et (Re)configuration de logiciels (DECOR'04), Grenoble (France), octobre 2004.
Continuous Measurement of Web Proxy Cache Efficiency. Simon Patarin et Mesaac Makpangou. Electronic Proceedings of the 12th International World Wide Web Conference (WWW03), Budapest, Hungary.
A Measurement System for E-Marketplaces. Ikram Chabbouh, Mesaac Makpangou and Hella Ben Ayed. The sixth International Conference on Electronic Commerce research, Dallas, Texas (USA), october 2003.
On-line Measurement of Web Proxy Cache Efficiency. Simon Patarin and Mesaac Makpangou. Inria Research Report, RR-4782, march 2003.
Pandora: A flexible network monitoring platform. Simon Patarin and Mesaac Makpangou. Usenix 2000 Annual Technical Conference, San Diego, june 2000.
Monitoring Web Proxy Caches. Simon Patarin et Mesaac Makpangou. INRIA Research Report RR-4023, INRIA, october 2000.
Saperlipopette! a Distributed Web Caching Systems Evaluation Tool. Guillaume Pierre and Mesaac Makpangou. Middleware Conference, september 1998.

Distributed Shared Objects

Adaptable Replicated Objects in Distributed Environments. Georges Brun-Cottan and Mesaac Makpangou. INRIA Research Report RR-2593, INRIA Rocquencourt, may 1995.
Fragmented Objects for Distributed Abstractions. Mesaac Makpangou, Yvon Gourhant, Jean-Pierre Le Narzul, and Marc Shapiro. In Readings in Distributed Computing Systems, pages 170 -- 186; Editors: T. L. Casavant and N. Singhal; IEEE Computer Society Press, july 1994.
A refinenment of the Fragmented Object Model. Peter Dickman and Mesaac Makpangou. The IEEE International WOrkshop on Object Orientation in Operating Systems, Dourdan (France), october 1992.
A Generic FO-structured Framework for Persistence Support in Distributed Settings. Hervé Soulard and Mesaac Makpangou. The IEEE International WOrkshop on Object Orientation in Operating Systems, Dourdan (France), october 1992.
Contrasting Fragmented Objects with Uniform Transparent Object References for Distribued Programming. Peter Dickman,Mesaac Makpangou and Marc Shapiro. The 5th European SIGOPS Workshop in Models and Paradigms for Distributed Systems Structuring, Mont Saint-Michel (France), september 1992. ACM SIGOPS.
Unification des paradigmes de partage en environnement réparti. Mesaac Makpangou et Georges Brun-Cottan. First African Conférence on Research in Computer Science, volume III, Yaoundé (Cameroon), october 1992.
BOAR: A Library of Fragmented Object Types for Distributed Abstractions. Mesaac Makpangou, Yvon Gourhant and Marc Shapiro. The International Workshop on Object-Orientation in Operating Systems. Palo Alto, CA(USA), october 1991.
Structuring distributed applications as fragmented objects. Mesaac Makpangou, Yvon Gourhant, Jean-Pierre Le Narzul, and Marc Shapiro. INRIA Research Report, RR-1404, INRIA Rocquencourt (France), january 1991.
Fragmented Object: a Building Block for Distributed Object-Support Operating Systems. Mesaac Makpangou and Yvon Gourhant. Newsletter of the Computer Society Technical Committee on Operating Systems and Application Environments, 5 (1):22--25, 1991.
The SOS Object-Oriented Communication Service. Mesaac Makpangou and Marc Shapiro. The 9th International Conference on Computer Communication, Tel Aviv (Israel) october 1988.
Invocations d'objets distants and SOS. Mesaac Makpangou. Conférence DNAC'88 : De nouvelles architectures pour les communications, Paris (France), octobre 1988.

Misc

A Flexible hybrid concurrency control model for collaborative applications in large scale settings. Guillaume Pierre and Mesaac Makpangou. The European ACM SIGOPS Workshop on Systems Support for Worldwide Applications, Connemara (Ireland), september 1996. ACM SIGOPS.
Designing application software in wide area network settings. Mesaac Makpangou and Ken Birman. Fourth European ACM SIGOPS Workshop on Fault-tolerance in Distributed Systems, Bologna (Italy), september 1990. ACM SIGOPS.