Mario Bravetti’s research while affiliated with University of Bologna and other places

Session types are widely used as abstractions of asynchronous message passing systems. Refinement for such abstractions is crucial as it allows improvements of a given component without compromising its compatibility with the rest of the system. In the context of session types, the most general notion of refinement is asynchronous session subtyping, which allows message emissions to be anticipated w.r.t. a bounded amount of message consumptions. In this paper we investigate the possibility to anticipate emissions w.r.t. an unbounded amount of consumptions: to this aim we propose to consider fair compliance over asynchronous session types and fair refinement as the relation that preserves it. This allows us to propose a novel variant of session subtyping that leverages the notion of controllability from service contract theory and that is a sound characterisation of fair refinement. In addition, we show that both fair refinement and our novel subtyping are undecidable. We also present a sound algorithm which deals with examples that feature potentially unbounded buffering. Finally, we present an implementation of our algorithm and an empirical evaluation of it on synthetic benchmarks.

We discuss an integrated approach for the design, specification, automatic deployment and simulation of microservice-based applications based on the ABS language. In particular, the integration of architectural modeling inspired by TOSCA (component types/port dependencies/architectural invariants) into the ABS language (static and dynamic aspects of ABS, including component properties, e.g., speed, and their use in timed/probabilistic simulations) via dedicated annotations. This is realized by the integration of the ABS toolchain with a dedicated tool, called Timed SmartDepl. Such a tool, at ABS code compile time, solves (starting from the provided architectural specification) the optimal deployment problem and produces ABS deployment orchestrations to be used in the context of timed simulations. Moreover, the potentialities and the expressive power of this approach are confirmed by further integration with external tools, e.g.: the Zephyrus tool, used by Timed SmartDepl to solve the optimal deployment problem via constraint solving, and a machine learning-based predictive module, that generates in advance data to be used in a timed ABS simulation exploiting such predicted data (e.g., simulating the usage, during the day, of predicted data generated during the preceding night).

In this work, we focus on by-design global scaling, a technique that, given a functional specification of a microservice architecture, orchestrates the scaling of all its components, avoiding cascading slowdowns typical of uncoordinated, mainstream autoscaling. State-of-the-art by-design global scaling adopts a reactive approach to traffic fluctuations, undergoing inefficiencies due to the reaction overhead. Here, we tackle this problem by proposing a proactive version of by-design global scaling able to anticipate future scaling actions. We provide four contributions in this direction: i) a platform able to host both reactive and proactive global scaling; ii) a proactive implementation based on data analytics; iii) a hybrid solution that mixes reactive and proactive scaling; iv) use cases and empirical benchmarks, obtained through our platform, that compare reactive, proactive, and hybrid global scaling performance. From our comparison, proactive global scaling consistently outperforms reactive, while the hybrid solution is the best-performing one.

Several emerging Industry 4.0 applications related to the monitoring and fault diagnostic of critical equipment introduce strict bounds on the latency of the data processing. Edge computing has emerged as a viable approach to mitigate the latency by offloading tasks to nodes nearby the data sources; at the same time, few industrial case studies have been reported so far. In this paper, we describe the design, implementation and evaluation of the SEAWALL platform for the heterogeneous data acquisition and low-latency processing in Industry 4.0 scenarios. The framework has been developed within the homonymous project founded by the Italian BIREX industrial consortium and involving both academic and industrial partners. The proposed framework supports data collection from heterogeneous production line machines mapped to different IoT protocols. In addition, it enables the seamless orchestration of workloads in the edge-cloud continuum so that the latency of the alerting service is minimized requirement of the processing task is continuously met, while taking into account the constrained resources of the edge servers. We evaluate the SEAWALL framework in a small-case industrial testbed and quantify the performance gain provided by the dynamic workload allocation on the continuum.

Detecting programming errors in software is increasingly important, and building tools that help developers with this task is a crucial area of investigation on which the industry depends. Leveraging on the observation that in Object-Oriented Programming (OOP) it is natural to define stateful objects where the safe use of methods depends on their internal state, we present Java Typestate Checker (JATYC), a tool that verifies Java source code with respect to typestates. A typestate defines the object's states, the methods that can be called in each state, and the states resulting from the calls. The tool statically verifies that when a Java program runs: sequences of method calls obey to object's protocols; objects' protocols are completed; null-pointer exceptions are not raised; subclasses' instances respect the protocol of their superclasses. To the best of our knowledge, this is the first OOP tool that simultaneously tackles all these aspects.

Session types are becoming popular and have been integrated in several mainstream programming languages. Nevertheless, while many programming languages consider asynchronous fifo channel communication, the notion of subtyping used in session type implementations is the one defined by Gay and Hole for synchronous communication. This might be because there are several notions of asynchronous session subtyping, these notions are usually undecidable, and only recently sound (but not complete) algorithmic characterizations for these subtypings have been proposed. But the fact that the definition of asynchronous session subtyping and the theory behind related algorithms are not easily accessible to non-experts may also prevent further integration. The aim of this paper, and of the tool presented therein, is to make the growing body of knowledge about asynchronous session subtyping more accessible, thus promoting its integration in practical applications of session types.

We develop a novel approach for run-time global adaptation of microservice applications, based on synthesis of architecture-level reconfiguration orchestrations. More precisely, we devise an algorithm for automatic reconfiguration that reaches a target system Maximum Computational Load by performing optimal deployment orchestrations. To conceive and simulate our approach, we introduce a novel integrated timed architectural modeling/execution language based on an extension of the actor-based object-oriented Abstract Behavioral Specification (ABS) language. In particular, we realize a timed extension of SmartDeployer, whose ABS code annotations make it possible to express architectural properties. Our Timed SmartDeployer tool fully integrates time features of ABS and architectural annotations by generating timed deployment orchestrations. We evaluate the applicability of our approach on a realistic microservice application taken from the literature: an Email Pipeline Processing System. We prove its effectiveness by simulating such an application and by comparing architecture-level reconfiguration with traditional local scaling techniques (which detect scaling needs and enact replications at the level of single microservices). Our comparison results show that our approach avoids cascading slowdowns and consequent increased message loss and latency, which affect traditional local scaling.