Luca De Martini
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Luca De Martini has a total of 11 co-authors. Here are the top co-authors by number of shared publications:
- Alessandro Margara
Politecnico di Milano, Milano, Italy
7 shared publications - Christoph Doblander
Technical University of Munich, Munich, Germany
2 shared publications - Giorgio Natale Buttiglieri
Dipartimento di Elettronica, Informazione e Bioingegneria (DEIB), Politecnico di Milano, Milano, Italy
2 shared publications - Jawad Tahir
Technical University of Munich, Munich, Germany
2 shared publications - Sebastian Frischbier
Allianz Global Investors, Frankfurt, Germany
2 shared publications Edoardo Morassutto
1 shared publication
Latest Publications
- Open Access
- 25 March 2026
- Published by Elsevier in Information Systems
- Vol. 140
Histrio: Actor-based programming and correctness guarantees for serverless environments
- 25 March 2026
- Published by Elsevier in Information Systems
- Vol. 140
Open AccessHistrio: Actor-based programming and correctness guarantees for serverless environments
- Open Access
- 9 June 2025
- Published by Association for Computing Machinery (ACM)
- p. 223-228
The DEBS 2025 Grand Challenge: Real-Time Monitoring of Defects in Laser Powder Bed Fusion (L-PBF) Manufacturing
- 9 June 2025
- Published by Association for Computing Machinery (ACM)
- p. 223-228
Open AccessThe DEBS 2025 Grand Challenge: Real-Time Monitoring of Defects in Laser Powder Bed Fusion (L-PBF) Manufacturing
- Open Access
- 9 June 2025
- Published by Association for Computing Machinery (ACM)
- p. 74-85
Histrio: a Serverless Actor System
- 9 June 2025
- Published by Association for Computing Machinery (ACM)
- p. 74-85
Open AccessHistrio: a Serverless Actor System
- Open Access
- 24 June 2024
- Published by Association for Computing Machinery (ACM)
- p. 223-228
The DEBS 2024 Grand Challenge: Telemetry Data for Hard Drive Failure Prediction
The DEBS Grand Challenge (GC) is a programming competition integrated into the annual ACM International Conference on Distributed and Event-Based Systems (DEBS) since DEBS 2011. The DEBS GC 2024 edition focuses on analyzing real-world telemetry data about over 200k hard drives in data centers. The goal of the challenge is to continuously compute clusters of similar drives as new data is received. This kind of analysis may be relevant to quickly identify groups of drives with a high probability of failure and enact predictive maintenance strategies. The submitted solutions are evaluated based on their design, performance, efficiency, ease of reuse, fault-tolerance, and adaptation to different problems. This paper describes the domain of telemetry data for hard drives, the data set used in the challenge, the specific problem formulation, and Challenger platform used to evaluate the submissions.
- 24 June 2024
- Published by Association for Computing Machinery (ACM)
- p. 223-228
Open AccessThe DEBS 2024 Grand Challenge: Telemetry Data for Hard Drive Failure Prediction
The DEBS Grand Challenge (GC) is a programming competition integrated into the annual ACM International Conference on Distributed and Event-Based Systems (DEBS) since DEBS 2011. The DEBS GC 2024 edition focuses on analyzing real-world telemetry data about over 200k hard drives in data centers. The goal of the challenge is to continuously compute clusters of similar drives as new data is received. This kind of analysis may be relevant to quickly identify groups of drives with a high probability of failure and enact predictive maintenance strategies. The submitted solutions are evaluated based on their design, performance, efficiency, ease of reuse, fault-tolerance, and adaptation to different problems. This paper describes the domain of telemetry data for hard drives, the data set used in the challenge, the specific problem formulation, and Challenger platform used to evaluate the submissions.
- Open Access
- 24 June 2024
- Published by Association for Computing Machinery (ACM)
- p. 30-41
Safe Shared State in Dataflow Systems
The complexity of implementing scalable data processing tasks in parallel and distributed computing environments has pushed the adoption of restricted programming models that simplify expressing such tasks. In this context, the dataflow model has emerged as the standard for building distributed data processing systems. The key insight of this model is to express computations as sequences of operations that do not share any state and can thus be deployed and executed independently on the same or on different machines. However, entirely excluding state sharing may be detrimental for performance, as it prevents operations executed on the same machine from accessing common resources. The effects of these limitations become more and more evident as the computational and memory resources of individual hosts increase, and even large-scale data analysis tasks can be performed with few well-equipped machines. Moving from this observation, in this paper we present an extension to the classic dataflow model that enables disciplined state sharing across operators to improve performance. We implement the model in Renoir, a dataflow system written in Rust, and we exploit the characteristics of this programming language to ensure safe access to shared state by design. We present several use cases where our model may be beneficial, and we use them to evaluate the proposed model in terms of performance and code complexity. We show that our model empowers developers to better exploit the resources of machines, resulting in substantial performance improvements in several use cases. Programs written with our extended model may even be more concise and simpler than in the classic dataflow model.
- 24 June 2024
- Published by Association for Computing Machinery (ACM)
- p. 30-41
Open AccessSafe Shared State in Dataflow Systems
The complexity of implementing scalable data processing tasks in parallel and distributed computing environments has pushed the adoption of restricted programming models that simplify expressing such tasks. In this context, the dataflow model has emerged as the standard for building distributed data processing systems. The key insight of this model is to express computations as sequences of operations that do not share any state and can thus be deployed and executed independently on the same or on different machines. However, entirely excluding state sharing may be detrimental for performance, as it prevents operations executed on the same machine from accessing common resources. The effects of these limitations become more and more evident as the computational and memory resources of individual hosts increase, and even large-scale data analysis tasks can be performed with few well-equipped machines. Moving from this observation, in this paper we present an extension to the classic dataflow model that enables disciplined state sharing across operators to improve performance. We implement the model in Renoir, a dataflow system written in Rust, and we exploit the characteristics of this programming language to ensure safe access to shared state by design. We present several use cases where our model may be beneficial, and we use them to evaluate the proposed model in terms of performance and code complexity. We show that our model empowers developers to better exploit the resources of machines, resulting in substantial performance improvements in several use cases. Programs written with our extended model may even be more concise and simpler than in the classic dataflow model.
- Open Access
- 13 June 2024
- Published by Elsevier in Future Generation Computer Systems
- Vol. 160, 472-488
The Renoir Dataflow Platform: Efficient Data Processing without Complexity
- 13 June 2024
- Published by Elsevier in Future Generation Computer Systems
- Vol. 160, 472-488
Open AccessThe Renoir Dataflow Platform: Efficient Data Processing without Complexity
- Open Access
- 11 December 2023
- Published by Association for Computing Machinery (ACM)
- p. 30-35
Nubes
Serveless computing and the Function-as-a-Service (FaaS) model promise rapid development of cloud-based applications by abstracting away deployment and resource allocation. As the stateless nature of functions undermines the generality of the model, they are often paired with storage services to persist their state. However, this approach exposes state management to developers, who need to manually encode the interactions between functions and storage. The relations between functions and state are hidden within function implementations, negatively affecting modularity and reuse. To overcome these problems, we propose a novel abstraction that brings the benefits of object-oriented programming to FaaS, and we implement this abstraction into the Nubes framework. In Nubes, developers define objects that encapsulate state in the form of attributes and expose methods to other objects. Applications are written using familiar object-oriented concepts, Nubes then transparently and automatically manages the state of objects using a cloud storage service and handles the execution of serverless functions. Nubes simplifies application development and deployment and promotes the reuse of objects as composable building blocks for cloud applications. Using a case study, we show that Nubes significantly reduces code complexity with limited overhead with respect to manually crafted solutions.
- 11 December 2023
- Published by Association for Computing Machinery (ACM)
- p. 30-35
Open AccessNubes
Serveless computing and the Function-as-a-Service (FaaS) model promise rapid development of cloud-based applications by abstracting away deployment and resource allocation. As the stateless nature of functions undermines the generality of the model, they are often paired with storage services to persist their state. However, this approach exposes state management to developers, who need to manually encode the interactions between functions and storage. The relations between functions and state are hidden within function implementations, negatively affecting modularity and reuse. To overcome these problems, we propose a novel abstraction that brings the benefits of object-oriented programming to FaaS, and we implement this abstraction into the Nubes framework. In Nubes, developers define objects that encapsulate state in the form of attributes and expose methods to other objects. Applications are written using familiar object-oriented concepts, Nubes then transparently and automatically manages the state of objects using a cloud storage service and handles the execution of serverless functions. Nubes simplifies application development and deployment and promotes the reuse of objects as composable building blocks for cloud applications. Using a case study, we show that Nubes significantly reduces code complexity with limited overhead with respect to manually crafted solutions.