2021/04/26 by Lorenzo Bacchiani, Mario Bravetti, Bacchiani, Lorenzo +9
Computer Science · #Advanced Software Engineering Methodologies #Cloud Computing and Resource Management #Distributed #FOS: Computer and information sciences #Parallel #Software System Performance and Reliability #and Cluster Computing (cs.DC)
paper · pdf · doi:10.48550/arxiv.2104.12466
openalex publication_date 2021/04/26 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
In the context of the BI-REX (Big Data Innovation and Research Excellence)\ncompetence center SEAWALL (SEAmless loW lAtency cLoud pLatforms) project\n(scientific coordinator Prof. Maurizio Gabbrielli) we develop a novel approach\nfor run-time global adaptation of microservice applications, based on synthesis\nof architecture-level reconfiguration orchestrations. More precisely, we devise\nan algorithm for automatic reconfiguration that reaches a target system Maximum\nComputational Load by performing optimal deployment orchestrations. To conceive\nand simulate our approach, we introduce a novel integrated timed architectural\nmodeling/execution language based on an extension of the actor-based\nobject-oriented Abstract Behavioral Specification (ABS) language. In\nparticular, we realize a timed extension of SmartDeployer, whose ABS code\nannotations make it possible to express architectural properties. Our Timed\nSmartDeployer tool fully integrates time features of ABS and architectural\nannotations by generating timed deployment orchestrations. We evaluate the\napplicability of our approach on a realistic microservice application taken\nfrom the literature: an Email Pipeline Processing System. We prove its\neffectiveness by simulating such an application and by comparing\narchitecture-level reconfiguration with traditional local scaling techniques\n(which detect scaling needs and enact replications at the level of single\nmicroservices). Our comparison results show that our approach avoids cascading\nslowdowns and consequent increased message loss and latency, which affect\ntraditional local scaling.\n