Distributed storage research at HASLab is motivated by the observation that storage systems are a cornerstone component for a wide range of research and industrial areas such as, cloud computing, high-performance computing, internet-of-things, databases, and artificial inteligence. Moreover, our research is driven by the need of such systems to provide efficient, scalable, resilient and secure solutions that can adapt to the exponential increase of digital information and to the different needs of very heterogeneous applications and infrastructures.
To this end, our research is focused on three main areas: Software-Defined Storage, Storage Benchmarking and Monitoring, Storage Optimization. In detail, with these three areas we intend to provide a new generation of programmable distributed storage systems that can adapt to heterogeneous storage workloads and deployment infrastructures, proper benchmarking tools for evaluating such systems, and novel storage functionalities and optimizations that can further improve the performance, reliability, and security of current solutions (see Research).
Our team is composed of researchers from the High-Assurance Software Laboratory (HASLab), mainly from the Distributed Systems field. HASLab is one of the integrated R&D centres of INESC TEC, a leading national associate laboratory, and a teaching and research subunit of the University of Minho, where it is headquartered.
Research at HASLab is anchored on a rigorous approach to three areas of Computer Science: Software Engineering, Distributed Systems and Cryptography and Information Security. The laboratory’s contributions to these areas range from fundamental research on formal methods and algorithms to applied research on tools and middleware that address real-world demands stemming from long-term collaborations with industry.
We got two papers -- BDUS: Implementing Block Devices in User Space and S2Dedup: SGX-enabled Secure Deduplication -- accepted at the SYSTOR'21 conference!10 October 2020
The PAStor project started on October 1st and will research new solutions to speedup storage performance for AI-oriented HPC services. We are very happy to be part of this collaboration between INESC TEC, TACC, MACC and Hood College.14 May 2020"
The paper entitled GenoDedup: Similarity-Based Deduplication and Delta-Encoding for Genome Sequencing Data, resulting from a collaboration with Vinicius Cogo and Alysson Bessani from LASIGE, was accepted in IEEE Transactions on Computers.20 April 2020
The BigHPC project kicked off on April 20 via videoconference. With $2M for 3 years, the project will develop new tools for improving the management of storage and computational resources at HPC infrastructures.30 March 2020
Our survey about software-defined storage was accepted in ACM CSUR17 November 2019
Ricardo Macedo earned a PhD grant from Fundação para a Ciência e Tecnologia!10 October 2019
We have published two new papers at the DRSS workshop colocated with SRDS'19!