Section: Partnerships and Cooperations

International Initiatives

Inria International Labs

DALHIS

Participants : Christine Morin, Deborah Agarwal, Anna Giannakou, Amir Teshome Wonjiga, Subarna Chatterjee.

• Title: Data Analysis on Large Heterogeneous Infrastructures for Science

• International Partner (Institution - Laboratory - Researcher):

  • Lawrence Berkeley National Laboratory (United States) - Data Science and Technology department - Deb Agarwal

• Start year: 2016

• See also: https://project.inria.fr/dalhis/

• Data produced by scientific instruments (large facilities like telescopes or field data), large-scale experiments, and high-fidelity simulations are increasing in magnitude and complexity. Existing data analysis methods, tools and infrastructure are often difficult to use and unable to provide the complete data management, collaboration, and curation environment needed to manage these complex, dynamic, and large-scale data analysis environments. The goal of the Inria-LBL DALHIS associate team involving the Myriads (PI) and Avalon Inria project-teams and the Data Science and Technology (DST) department at Lawrence Berkeley National Laboratory (LBL) is to create a collaborative distributed software ecosystem to manage data lifecycle and enable data analytics on distributed data sets and resources. Specifically, our goal is to build a dynamic software stack that is user-friendly, scalable, energy-efficient and fault tolerant. Our research will determine appropriate execution environments that allow users to seamlessly execute their end-to-end dynamic data analysis workflows in various resource environments and scales while meeting energy-efficiency, performance and fault tolerance goals. We will engage in deep partnerships with scientific teams (Fluxnet in environmental science and SNFactory and LSST experiences in cosmology) and use a mix of user research with system software R&D to address specific challenges that these communities face. In 2017, we worked on evaluating data streaming environments (see Section 7.1.4) and on producing tools to help users (scientist in the climate and environment community) to explore the carbon flux datasets from AmeriFlux (Americas) and FLUXNET (global) (see Section 7.4.2.2). We also worked on two facets of security in the context of HPC distributed computing infrastructures: (i) building a workflow for data analysis for anomaly detection and (ii) using the block-chain technology to leverage data integrity at the network and some portion of computation levels.

Inria Associate Teams Not Involved in an Inria International Labs

SUSTAM associated team

Participants : Anne-Cécile Orgerie, Yunbo Li.

Anne-Cécile Orgerie participates in the associated team named SUSTAM (Sustainable Ultra Scale compuTing, dAta and energy Management) leaded by Laurent Lefévre (Avalon team, Lyon) with Prof. Manish Parashar (RDI2, Rutgers University, NJ, USA). The SUSTAM associate team will focus on the joint design of a multi-criteria orchestration framework dealing with resources, data and energy management in an sustainable way.

Inria International Partners

Informal International Partners

• Partner: Rutgers University, dept. of Computer Science (New Jersey, United States)

• We collaborate with Manish Parashar's research group on energy efficiency in edge Clouds and in particular on the design of energy cost models for such environments involving renewable energy. A joint paper has been presented at IEEE/ACM CCGrid 2017.

• Partner: Northeastern University, dept. of Computer Science (Massachusetts, United States)

• We collaborate with Gene Cooperman 's research group on the study of large-scale distributed systems. More specifically, we actively collaborate on virtualization technologies and system snapshoting (we obtained a postdoc funding on that topic from the Brittany Regional Council, but all applicants declined in the last minute). We plan to reinforce and extend our collaboration to formal methods for distributed systems in the next year.

• Partner: University of Guadalajara (Mexico)

• We collaborate with the team of Prof. Hector Duran-Limon on application and resource management in the cloud. In 2017, we produced a joint publication [46]. Nikos Parlavantzas is co-advising a PhD student enrolled in the University of Guadalajara (Carlos Ruiz Diaz).

• Partner: Tlemcen University (Algeria)

• We collaborate with Djawida Dib on energy-efficient fault-tolerant resource and application management in containerized clouds. Christine Morin and Nikos Parlavantzas have been co-advising Yasmina Bouizem, a PhD student enrolled in the University of Tlemcen from December 2016.