Rolando Martins | Carnegie Mellon University (original) (raw)
Rolando Martins studied at Faculty of Science of the University ofPorto (FCUP), where he also obtained his M.Sc in Informatics:Networks and Systems. As part of his Masters thesis (YapDss),he researched the field of distributed stack splitting in Prolog,exploring Or-Parallelism. He also worked at EFACEC as a softwareengineer/architect and later as a systems researcher.He obtained his Ph.D in Computer Science from FCUP, as a part of acollaborative effort between FCUP, EFACEC and Carnegie MellonUniversity, under the supervision of Prof. Fernando Silva, Prof. LuísLopes and Prof. Priya Narasimhan.His Ph.D. research topic arose from his employment at EFACEC, where hewas exposed to the difficulties underlying today’s railway systems andlight-rail deployments, and came to understand the scientificchallenges and the impact, of addressing the issues of simultaneouslysupporting real-time and fault-tolerance in such systems.He is a former member of the the Intel Science and Technology Center(ISTC), where he was involved in both Cloud Computing and EmbeddedComputing centers, and Parallel Data Lab (PDL) at Carnegie MellonUniversity (CMU). At the same time, he was also a computer researchscientist at YinZcam, a spinoff from CMU that provided mobileapplications for the NBA, NHL, NFL and MLS, where he was involved oncloud computing, content management systems, OAuth and videostreaming. He is currently an invited assistant professor at the department ofComputer Science at FCUP and researcher at CRACS (Center for Researchin Advanced Computing Systems) part of INESC-Tec. Some of his researchinterests include security, privacy, intrusion tolerance, (secure)distributed systems, edge clouds, P2P, IoT, cloud-computing,fault-tolerance (byzantine and non-byzantine), operating systems (withspecial interest in the Linux kernel).
less
Uploads
Papers by Rolando Martins
We address the problem of integrating real-time fault-tolerance mechanisms into peer-to-peer syst... more We address the problem of integrating real-time fault-tolerance mechanisms into peer-to-peer systems, with specific architecture and deployment constraints. For this purpose we implemented a prototype hierarchical peer-to-peer framework in which the leaf peers are sensors that generate different kinds of traffic such as mesh management, events, video and audio. We evaluate the framework by measuring packet loss, response time, jitter and mesh overhead for each type of traffic under peer failure. We report significant gains in QoS for all types of traffic, when using the fault-tolerance mechanisms, with minimal response time and mesh management overhead.
We address the problem of providing transparent, lightweight, fault-tolerance mechanisms for gene... more We address the problem of providing transparent, lightweight, fault-tolerance mechanisms for generic peerto-peer middleware systems. The main idea is to use the peer-to-peer overlay to provide for fault-tolerance rather than support it higher up in the middleware architecture, e.g. in the form of services. To evaluate our approach we have implemented a fault-tolerant middleware prototype that uses a hierarchical peer-to-peer overlay in which the leaf peers connect to sensors that provide data streams. Clients connect to the root of the overlay and request streams that are routed upwards through intermediate peers in the overlay up to the client. We report encouraging preliminary results for latency, jitter and resource consumption for both the non-faulty and faulty cases.
In this paper we present the architecture of RTPM , a middleware framework aimed at supporting th... more In this paper we present the architecture of RTPM , a middleware framework aimed at supporting the development and management of information systems for high-speed public transportation systems. The framework is based on a peerto-peer overlay infrastructure with the main focus being on providing a scalable, resilient, reconfigurable, highly available platform for real-time and QoS computing.
This paper discusses the design of YapDss, an or-parallel Prolog system for distributed memory pa... more This paper discusses the design of YapDss, an or-parallel Prolog system for distributed memory parallel machines, such as the Beowulf PC clusters. The system builds on the work of YapOr, an or-parallel system for shared memory machines, and uses the distributed stack splitting binding model to represent computation state and work sharing among the computational workers. A new variant scheme of stack splitting, the diagonal splitting, is proposed and implemented. This scheme includes efficient algorithms to balance work load among computing workers, to determine the bottommost common node between two workers, and to calculate exactly the work load of one worker. An initial evaluation of the system shows that it is able to achieve very good speedups on a Beowulf PC cluster.
We address the problem of integrating real-time fault-tolerance mechanisms into peer-to-peer syst... more We address the problem of integrating real-time fault-tolerance mechanisms into peer-to-peer systems, with specific architecture and deployment constraints. For this purpose we implemented a prototype hierarchical peer-to-peer framework in which the leaf peers are sensors that generate different kinds of traffic such as mesh management, events, video and audio. We evaluate the framework by measuring packet loss, response time, jitter and mesh overhead for each type of traffic under peer failure. We report significant gains in QoS for all types of traffic, when using the fault-tolerance mechanisms, with minimal response time and mesh management overhead.
We address the problem of providing transparent, lightweight, fault-tolerance mechanisms for gene... more We address the problem of providing transparent, lightweight, fault-tolerance mechanisms for generic peerto-peer middleware systems. The main idea is to use the peer-to-peer overlay to provide for fault-tolerance rather than support it higher up in the middleware architecture, e.g. in the form of services. To evaluate our approach we have implemented a fault-tolerant middleware prototype that uses a hierarchical peer-to-peer overlay in which the leaf peers connect to sensors that provide data streams. Clients connect to the root of the overlay and request streams that are routed upwards through intermediate peers in the overlay up to the client. We report encouraging preliminary results for latency, jitter and resource consumption for both the non-faulty and faulty cases.
In this paper we present the architecture of RTPM , a middleware framework aimed at supporting th... more In this paper we present the architecture of RTPM , a middleware framework aimed at supporting the development and management of information systems for high-speed public transportation systems. The framework is based on a peerto-peer overlay infrastructure with the main focus being on providing a scalable, resilient, reconfigurable, highly available platform for real-time and QoS computing.
This paper discusses the design of YapDss, an or-parallel Prolog system for distributed memory pa... more This paper discusses the design of YapDss, an or-parallel Prolog system for distributed memory parallel machines, such as the Beowulf PC clusters. The system builds on the work of YapOr, an or-parallel system for shared memory machines, and uses the distributed stack splitting binding model to represent computation state and work sharing among the computational workers. A new variant scheme of stack splitting, the diagonal splitting, is proposed and implemented. This scheme includes efficient algorithms to balance work load among computing workers, to determine the bottommost common node between two workers, and to calculate exactly the work load of one worker. An initial evaluation of the system shows that it is able to achieve very good speedups on a Beowulf PC cluster.