.png){kind=icon}
.png){kind=icon}
Return
Workflow Refactoring for Maximizing Concurrency and Block-Structuredness
.png){kind=icon}
PRE
AI
DOI：10.1109/TSC.2018.2867593.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
In the era of Internet and big data, contemporary workflows become increasingly large in scale and complex in structure, introducing greater challenges for workflow modeling. Workflows are not with maximized concurrency and block-structuredness in terms of control flow, though languages supporting block-structuredness (e.g., BPEL) are employed. Existing workflow refactoring approaches mostly focus on maximizing concurrency according to dependences between activities, but do not consider the block-structuredness of the refactored workflow. It is easier to comprehend and analyze a workflow that is block-structured and to transform it into BPEL-like processes. In this paper, we aim at maximizing both concurrency and block-structuredness. Nevertheless, not all workflows can be refactored with a block-structured representation, and it is intractable to make sure that the refactored workflows are as block-structured as possible. We first define a well-formed dependence pattern of activities. The control flow among the activities in this pattern can be represented in block-structured forms with maximized concurrency. Then, we propose a greedy heuristics-based graph reduction approach to recursively find such patterns. In this way, the resulting workflow is with maximized concurrency and its block-structuredness approximates optimality. We show the effectiveness and efficiency of our approach with real-world scientific workflows.
Keywords:
Concurrent computing
Transforms
Synchronization
Electronic mail
Data models
Collaboration
Optimal control
Workflow refactoring
activity dependence
concurrency maximization
block-structuredness
synchronization links
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
5.8
Papers:
2.1K
Citations:
6.5K