Refactoring Java Code to MapReduce Framework (S)
Junfeng Zhao · Proceedings/Proceedings of the ... International Conference on Software Engineering and Knowledge Engineering · 2021
Not cached yet. This record was fetched live for this request and is stored after the lookup. Reopening the page should serve the cached copy.
Proceedings/Proceedings of the ... International Conference on Software Engineering and Knowledge Engineering · OpenAlex · conference-paper · 2021
Abstract
Cloud computing has evolved into an infrastructure tool for scientific research and computing application.For many enterprises, it has become a trend to migrate their applications from the local to cloud.To leverage cloud computing infrastructure, some legacy code for special business process need to be refactored to the programming models of cloud computing.The desired approach is to design an automatic tool for refactoring legacy code to the target code that can execute on cloud computing platform effectively.In this paper a new approach is proposed, which can automatically refactor Java sequential program into MapReduce paradigm.The approach works by first translating input code into functional representation, with loops succinctly encapsulated by fold operations.Then, guided by the transforming rules, the approach generates equivalent MapReduce programs.The rules use group-by operations to enable greater parallelism.Finally, a series of mapping rules is applied to map immediate code to the target code running on Spark.A new tool designed using this approach, JMRT, was evaluated using real world benchmarks.The experimental results show that the approach can generate the desired MapReduce program and the business execution efficiency can be improved.
Open the resolved locator · 10.18293/seke2021-059
Metadata licence: OpenAlex metadata (CC0)