A Pareto-Efficient Algorithm for Data Stream Processing at Network Edges

Loukopoulos, Thanasis; Tziritas, Nikos; Koziri, Maria; Stamoulis, Georgios; Khan, Samee U.

doi:10.1109/cloudcom2018.2018.00041

Cited by 7 publications

(2 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Silva et al [23] propose a methodology for monitoring communication metrics on the edges of DAGs as the training set of reinforcement learning to reconfigure the topology. Moreover, previous studies [24][25][26] propose different task scheduling strategies for minimizing communication overhead and reducing processing latency. Since data transition contributes the most overhead to the cluster, they reschedule tasks with the most data transitions to the same computing node to reduce communication overhead across nodes, which is an effective method for improving the efficiency of data transition and processing.…”

Section: Related Workmentioning

confidence: 99%

Flink-ER: An Elastic Resource-Scheduling Strategy for Processing Fluctuating Mobile Stream Data on Flink

Bian

et al. 2020

Mobile Information Systems

View full text Add to dashboard Cite

As real-time and immediate feedback becomes increasingly important in tasks related to mobile information, big data stream processing systems are increasingly applied to process massive amounts of mobile data. However, when processing a drastically fluctuating mobile data stream, the lack of an elastic resource-scheduling strategy limits the elasticity and scalability of data stream processing systems. To address this problem, this paper builds a flow-network model, a resource allocation model, and a data redistribution model as the foundation for proposing Flink with an elastic resource-scheduling strategy (Flink-ER), which consists of a capacity detection algorithm, an elastic resource reallocation algorithm, and a data redistribution algorithm. The strategy improves the performance of the platform by dynamically rescaling the cluster and increasing the parallelism of operators based on the processing load. The experimental results show that the throughput of a cluster was promoted under the premise of meeting latency constraints, which verifies the efficiency of the strategy.

show abstract

Section: Related Workmentioning

confidence: 99%

Flink-ER: An Elastic Resource-Scheduling Strategy for Processing Fluctuating Mobile Stream Data on Flink

Bian

et al. 2020

Mobile Information Systems

View full text Add to dashboard Cite

show abstract

“…These services are charged accordingly to the amount and duration of used computing capabilities. In this scenario, some works have emerged aiming at the reduction of the cost involved using these services (Azumah et al, 2018;Lin et al, 2017;Loukopoulos et al, 2018). They tackle this problem by establishing a minimal task set that still produces the desired results but requires minimal computing capabilities as possible.…”

Section: Related Workmentioning

confidence: 99%

Hybrid Cloud Rendering for Industrial-Plant Cad Models

MOREIRA¹

View full text Add to dashboard Cite

Industrial-plant CAD models play an important role in engineering project management. Despite the advances in computing power in past decades, rendering these models remains challenging due to their complexity and large data volume. Different areas of computing have succeeded in adopting cloud services to process massive data. However, when it comes to cloud rendering, there is still a lack of cloud rendering services for CAD models. In this paper, we propose a hybrid cloud rendering architecture for CAD models, dividing the rendering task between client and server. In addition to reducing server overhead, this approach affords greater resilience to the system against variations of network latency. Finally, this work also introduces a metaheuristic-based workload selection algorithm to determine the set of objects to be drawn on the client side. Our results demonstrate that the proposed methodology allows efficient visualization of massive CAD models even under adverse conditions such as clients with limited devices and high connection latency. Lastly, we discuss remaining research opportunities for cloud rendering, opening avenues for future improvements.

show abstract