Workloads Scheduling in Scientific Workflow-as-a-Service Platforms

This chapter delves into the complexities of workload scheduling in scientific Workflow-as-a-Service (WaaS) platforms, highlighting the challenges and solutions associated with managing and optimizing cloud resources. It begins by introducing the Critical Jobs' Method (CJM) and its modifications to handle dynamic workflows efficiently. The text then explores various virtual resource allocation strategies, including greedy, control, and mixed strategies, each with its own advantages and trade-offs. A significant focus is placed on the Follow The Leader (FTL) algorithm for grouping tasks into parallel batches and the VM Allocation (VMA) algorithm for dynamic VM allocation, both of which are crucial for optimizing resource usage. The chapter also presents software implementation details and optimization results for real-world workflows like LIGO and MONTAGE, demonstrating the practical applicability of the proposed methods. By the end, readers will gain insights into advanced scheduling techniques and their impact on cloud resource management, making it a valuable resource for professionals seeking to enhance their understanding of workflow optimization in cloud computing environments.