Anthony Dugois

PhD in Computer Science

Find me on ORCID, Google Scholar, dblp, GitHub and LinkedIn.

Say hello at anthony.dugois@univ-fcomte.fr or download my CV (french or english).

Bonjour ! đź‘‹

I am a Teaching and Research Assistant (Attaché Temporaire d'Enseignement et de Recherche) at Université de Franche-Comté, in the marvelous city of Besançon. I teach various courses in Computer Science, such as web programming, networks, software testing, as well as compilers and interpreters.

I was previously a PhD student at École Normale Supérieure de Lyon, where I worked under supervision of Loris Marchal and Louis-Claude Canon in the Laboratoire de l'Informatique du Parallélisme.

My research interests include scheduling algorithms, optimization in parallel computing and distributed storage systems. I have also a keen interest in approximation theory.

Research

I study scheduling problems arising in distributed, replicated and persistent storage systems. The objective is to derive guarantees on response time and load balancing of these systems in an online context, that is to say, without any prior knowledge of the workload.

More recently, I started working on alternative analysis techniques for online scheduling problems, such as learning-augmented scheduling or parameterized worst-case analysis.

Recent Talks

  • ICPP 2023, Salt Lake City, Utah (August 2023): “Hector: A Framework to Design and Evaluate Scheduling Strategies in Persistent Key-Value Stores” (slides)
  • IPDPS 2022, visio-confĂ©rence (June 2022): “Bounding the Flow Time in Online Scheduling under Structured Processing Sets” (slides)
  • Euro-Par 2021, visio-confĂ©rence (September 2021): “Taming Tail-Latency in Key-Value Stores: a Scheduling Perspective” (slides)

Scientific Publications

Authors appear in alphabetical order.

  • Under reviewCanon, L.-C., Dugois, A., El Sayah, M., & HĂ©am, P.-C. (2024). A Markov Chain Monte Carlo Approach to Cost Matrix Generation for Scheduling Performance Evaluation. In Future Generation Computer Systems.
  • Under reviewCanon, L.-C., Dugois, A., & Marchal, L. (2024). Solving the Restricted Assignment Problem to Schedule Multi-Get Requests in Key-Value Stores. In Euro-Par 2024.
  • NewBen Mokhtar, S., Canon, L.-C., Dugois, A., Marchal, L., & Rivière, E. (2024). A Scheduling Framework for Distributed Key-Value Stores and its Application to Tail Latency Minimization. In Journal of Scheduling. DOI. PDF.
  • Dugois, A. (2023). Scheduling in Distributed Storage Systems. PDF.
  • Canon, L.-C., Dugois, A., Marchal, L., & Rivière, E. (2023). Hector: A Framework to Design and Evaluate Scheduling Strategies in Persistent Key-Value Stores. In ICPP 2023 (pp. 535-545). DOI. PDF.
  • Canon, L.-C., Dugois, A., & Marchal, L. (2022). Bounding the Flow Time in Online Scheduling with Structured Processing Sets. In IPDPS 2022 (pp. 683-693). DOI. PDF.
  • Canon, L.-C., Dugois, A., & Marchal, L. (2022). Bounding the Flow Time in Online Scheduling with Structured Processing Sets (extended version). PDF.
  • Ben Mokhtar, S., Canon, L.-C., Dugois, A., Marchal, L., & Rivière, E. (2021). Taming Tail Latency in Key-Value Stores: a Scheduling Perspective. In Euro-Par 2021 (pp. 136-150). DOI. PDF.
  • Ben Mokhtar, S., Canon, L.-C., Dugois, A., Marchal, L., & Rivière, E. (2021). Taming Tail Latency in Key-Value Stores: a Scheduling Perspective (extended version). PDF.

Teaching

In addition to my research activities, I work as a teaching assistant. Covered topics include parallel algorithms, distributed programs, software testing, networks, programming, and compilers. Courses are taught in French (fr) or English (en).

YearTopicTypeLevelLocation
2023-2024Bases de la Programmation (fr)TD/TPL1UFC
Langages du Web (fr)TPL2UFC
Web Avancé (fr)TPL3UFC
Analyse Syntaxique (fr)TPL3UFC
Encadrement Projet & Stage (fr)—L3UFC
RĂ©seaux (fr)TPM1UFC
Compilation & Interprétation (fr)TPM1UFC
Test Fonctionnel (fr)TPM2UFC
2022-2023Bases de la Programmation (fr)TD/TPL1UFC
RĂ©seaux (fr)TPM1UFC
2021-2022Circuits Logiques et RĂ©seaux (fr)TD/TPL3ENSL
Parallel and Distributed Algorithms & Programs (en)TD/TPM1ENSL
2020-2021Architecture, Système et Réseaux (fr)TD/TPL3ENSL
Integrated Project (en)TD/TPM1ENSL

Software

In an effort to make my experiments fully reproducible, all pieces of software used to generate the results of my publications are available and open.

  • Hector
    A framework built on top of Apache Cassandra to design, implement and evaluate scheduling algorithms in an industry-standard system.
  • Hector Benchmarking
    Experiments performed in the ICPP paper entitled "Hector: A Framework to Design and Evaluate Scheduling Strategies in Persistent Key-Value Stores".
  • Theoretical Throughput Estimator
    Experiments performed in the IPDPS paper entitled "Bounding the Flow Time in Online Scheduling with Structured Processing Sets".
  • Key-Value Store Simulator
    Experiments performed in the Euro-Par paper entitled "Taming Tail Latency in Key-Value Stores: a Scheduling Perspective".

Interests

In my spare time, I volunteer with the Climate Fresk to help raising public awareness of the mechanisms that are at the root of climate change.

I also enjoy experimenting with various recipes in my kitchen (who doesn't love food?) and relaxing in nature during long running sessions.