Alexandru Costan

Associate Professor, IRISA / INSA Rennes

 

IRISA / Inria Rennes – Bretagne Atlantique
Office: D169
Campus Universitaire de Beaulieu
35042, Rennes, France

Phone: + 33 2  99 84 75 66

E-mail: alexandru.costan at irisa.fr

 

Research interests

My main research interests are: Big Data management on large scale infrastructures like clouds, fog, edge, stream processing and workflow data management.

Recent highlights


Short bio

Alexandru Costan is an Associate Professor at INSA Rennes and a researcher within the KerData team at IRISA Rennes. In 2011, he obtained a Ph.D. in Computer Science from the Politehnica University of Bucharest (PUB), under the supervision of Valentin Cristea (PUB) and Iosif Legrand (California Institute of Technology). The Ph.D. thesis focused on self-adaptive behavior of large-scale distributed systems based on monitoring information, bringing several contributions to the MonALISA monitoring system, developed in collaboration with Caltech and CERN. After the Ph.D., Alexandru joined Inria as a postdoctoral researcher within the KerData team, working with Gabriel Antoniu and Luc Bougé, on scalable storage in cloud environments. In 2012, he became an Associate Professor at INSA Rennes, where he is currently leading the Big Data Science track. His research interests include Big Data management in HPC and clouds, fast data and stream processing, autonomic behavior and workflow management. At IRISA, he is currently involved in several research projects working on efficient Big and Fast Data management on clouds and edge infrastructures in the context of several collaborations with Microsoft Research and Huawei Research. Alexandru has published one book, more than 20 articles in international journals and 30 papers in international conferences. He serves as PC member of several top-level conferences and workshops in the domain of distributed computing (SuperComputing, CCGrid, Cluster, Big Data). Since 2011 he is the co-chair of the BigDataCloud workshop at EuroPar as well as the ScienceCloud workshop at HPDC (since 2015). He is currently leading the ANR OverFlow project and and he is a member of the JLESC: Joint Laboratory on Extreme-Scale Computing.


Publications

A list of my recent publications can be found on HAL Open Archives Library and DBLP:


Software

  • Planner is middleware for cost-efficient execution plans placement for uniform stream analytics on Edge and Cloud. Planner automatically selects which parts of the execution graph will be executed at the Edge in order to minimize the network cost. Real-world micro-benchmarks show that Planner reduces the network usage by 40% and the makespan (end-to-end processing time) by 15% compared to state-of-the-art.
  • KerA is a low-latency storage for stream processing (currently under development at Inria, in collaboration with Universidad Politécnica de Madrid, in the framework of a contractual partnership between Inria and Huawei Munich). By eliminating storage redundancies between data ingestion and storage, preliminary experiments with KerA successfully demonstrated its capability to increase throughput for stream processing.
  • Tyr is a transactional object storage system aimed at storage-based convergence between HPC and Big Data. Tyr natively offers data access coordination in the form of transactions. It offers a POSIX-compliant, transactional, distributed file system implementation built as a thin layer atop. This file system performs well and shows near-linear scalability properties on both HPC and Big Data platforms.
  • JetStream is a high performance batch-based streaming middleware. The system is able to self-adapt to the streaming conditions by modeling and monitoring a set of context parameters. It further aggregates the available bandwidth by enabling multi-route streaming across cloud sites.
  • TomusBlobs is a concurrency-optimized data storage system which federates the virtual disks associated to the Virtual Machines running the application code on the cloud. It is used in the Azure cloud for efficient data-intensive processing based on MapReduce.
  • MonALISA, which stands for Monitoring Agents using a Large Integrated Services Architecture, has been developed over the last years by Caltech and its partners with the support of the U.S. CMS software and computing program. The framework is based on Dynamic Distributed Service Architecture and is able to provide complete monitoring, control and global optimization services for complex systems.

Projects

  • ANR OverFlow (2015-2019): I am the Principal Investigator of this ANR JCJC project focusing on Workflow Data Management as a Service for Multi-site Applications. Budget: 250,000 Euros.
  • BigStorage (2015-2018) is an European Training Network (ETN) project. Area: Storage-based Convergence of HPC and Cloud infrastructures to handle Big Data. Role: local coordinator for Inria Rennes Bretagne Atlantique.
  • Z-CloudFlow (2013-2016): geographically distributed workflows on Azure clouds. A project of the Microsoft-Inria Joint Research Centre.

Postdocs

  • Pedro Silva (2018-2019). Research topic: “Stream processing on Cloud, Edge and hybrid Cloud/Edge environments”

PhD students

  • Paul Le Noac’h (2016-2019) co-advised with Luc Bougé (ENS Rennes / IRISA). Topic: “Workflow Data Management as a Service for Multi-Site Applications”

Former postodcs


Former PhD students

  • Ovidiu Marcu (2015-2018), co-advised with Gabriel Antoniu (Inria) and Maria S. Péréz (Universidad Politécnica de Madrid). Topic: “KerA: A Unified Ingestion and Storage System for Scalable Big Data Processing” . Currently, Research Engineer at Inria Rennes, France
  • Pierre Matri (2015-2018), co-advised with Gabriel Antoniu (Inria) and Maria S. Péréz (Universidad Politécnica de Madrid). Topic: “Tyr: Storage-Based HPC and Big Data Convergence Using Transactional Blob”. Currently, Postdoctoral Researcher at Argonne National Laboratory, USA
  • Luis Pineda (2014-2017), co-advised with Gabriel Antoniu (Inria). Topic: “Efficient support for data-intensive scientific workflows on geo-distributed clouds”. Currently, Research Engineer at Activeeon 
  • Radu Tudoran (2011-2014), co-advised with Gabriel Antoniu (Inria) and Luc Bougé (ENS Rennes / IRISA). Topic: “Efficient Big Data Management across Cloud Data Centers”. Currently, Principal Research Engineer at  Huawei European Research Center, Germany

Teaching