Education

Ph.D. degree in Computer Science, Technical University of Catalonia, 2008.
M.S. degree in Bioinformatics and Biostatistics, Universitat Oberta de Catalunya, 2018.
M.S. degree in Computer Science, Technical University of Catalonia, 2004.
 

About me

My research career has focused on two primary areas, corresponding to the two distinct phases of my career. During my predoctoral training, I concentrated on resource orchestration, primarily employing simulation techniques and heterogeneous clusters. My principal goal was to develop efficient models using simulation to address server saturation issues. This was accomplished by managing the various applications running on the systems and successfully reducing the service level agreements (SLAs) required by these applications.

Upon completion of my doctoral studies, I assumed the role of team and project manager within the Storage Systems group at the Barcelona Supercomputing Center, during the absence of the group leader due to parental leave. I got the roles of team management and project management (XtreemOS project), still related to orchestration and scheduling, and started my Storage research in the next months.

My storage research has been dedicated to 1) optimising the system I/O, as occurred in the IOLanes project and IOStack, with work on the Linux I/O Schedulers (automatic optimization of the I/O scheduler using pattern matching and prediction of the workload), work on Openstack Swift (Optimization to have a better interface with the kernel), 2) Enhancing the performance of storage, with contributions to the optimization of reliability methods and the use of energy for storage systems, and 3) Ephemeral filesystems, with the research of adhoc-fs, GekkoFS being one of the key achievements. GekkoFS achieved greater performance than other Parallel File Systems (GPFS or Lustre) using local resources as a building block for a distributed file system.

GekkoFS is oriented to HPC workloads, and as future objectives I would investigate its usage and benefits on serverless environments and machine learning environments. Also I have several open collaborations with other teams at Barcelona Supercomputing Center to study its usage in Life Science and Earth Science. Finally, I am participating in a RISC-V initiative and ported GekkoFS to RISC-V, that will open several new collaborations in the future due EU interest on an European processor.

Leadership

Actually I am Group Leader of the Data Centric Computing Team, whose research interests closely align with my PhD specialisation in orchestration. I am resuming my contributions in orchestration, so I will address the team's gap in I/O resource orchestration as I have done recently in the NEARDATA project. Additionally, given that machine learning is a key research focus of the team, I intend to leverage it in two primary directions. Firstly, I aim to explore the integration of ephemeral filesystems with machine learning to enhance I/O. Secondly, I plan to utilise machine learning within I/O systems to predict and orchestrate workloads effectively. I am actively doing it on a daily basis as I am participating in all the related activities of the different projects of the team : CLOUDSTARS, NEARDATA, VITAMIN-V and SECURED. Also one of my tasks is to ensure that there is financement on the future, so I am looking forward actively to new projects