A general-purpose distributed computing Java middleware

摘要

The middleware solutions for General-Purpose Distributed Computing (GPDC) have distinctrequirements, such as task scheduling, processing/storage fault tolerance, code portability forparallel or distributed environments, simple deployment (including over grid ormulti-cluster environments),collaborative development, low code refactoring, native support for distributed datastructures, asynchronous task execution, and support for distributed global variables. These solutionsdo not integrate these requirements into a single deployment with a unique API exposingmost of these requirements to users. The consequence is the utilization of several solutions withtheir particularities, thus requiring different user skills. Besides that, the users have to solve theintegration and all heterogeneity issues. To reduce this integration gap, in this paper, we presentJava Cá&Lá (JCL), a distributed-shared-memory and task-oriented lightweight middleware forthe Java community that separates business logic from distribution issues during the developmentprocess and incorporates several requirements that were presented separately in theGPDCmiddleware literature over the last few decades. JCL allows building distributed or parallel applicationswith only a few portable API calls, thus reducing the integration problems. Finally, it alsoruns on different platforms, including small single-board computers. This work compares andcontrasts JCL with other Java middleware systems and reports experimental evaluations of JCLapplications in several distinct scenarios.