Affinity-aware resource provisioning for long-running applications in shared clusters

Mommessin, Clément; Yang, Renyu; Shakhlevich, Natalia V.; Sun, Xiaoyang; Kumar, Satish; Xiao, Junqing; Xu, Jie

doi:10.1016/j.jpdc.2023.02.011

Cited by 4 publications

(2 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We categorize edge nodes into bandwidth provider and bandwidth receivers, according to the provision/shortage role, i.e., negative/positive value within the attribute tuple. Our previous algorithmic study on bin packing [28] concluded that Given that node-centric algorithms can achieve a good trade-off between running time and solution quality. Based on this, we propose an approximation Redirection mechanism, Prioritized Max-Request First Served (PMRFS) algorithm, to i) prioritize the bandwidth request from the high-priority queue with a larger amount of bandwidth shortage; and ii) consider edge nodes that has larger available bandwidth as the bandwidth providers.…”

Section: Prioritized Max Request First Served Heuristicmentioning

confidence: 99%

“…We are aware of many other counterpart heuristics that can serve the requests of traffic redirection. Specifically, our previous study [28] presented comprehensive experimental comparisons among node-centric, application (request)centric and multi-node approaches. As will be shown in §7.1, we select the following representative winning algorithms as baselines to compare their runtime performance.…”

Section: Other Comparative Heuristicsmentioning

confidence: 99%

See 1 more Smart Citation

Janus: Latency-Aware Traffic Scheduling for IoT Data Streaming in Edge Environments

Wen,

Yang,

Qian

et al. 2023

IEEE Trans. Serv. Comput.

Self Cite

View full text Add to dashboard Cite

This paper focuses on a simple, yet fundamental question of distributed edge computing: "how to handle IoT traffic with different levels of sensitivity and criticality by satisfying the application-specific latency constraints?" This question arises in the practical deployment of edge computing, where user data can arrive at a much faster rate than that they can be processed by an edge node. Addressing this question is critical for meeting the latency requirement for latency-sensitive applications, but existing approaches are inadequate to the problem. We present JANUS, a multi-level traffic scheduling system for managing multiple data streams with various degrees of latency constraints. At the edge node level, JANUS uses multi-level queues to manage data streams with different latency constraints. It then allocates the output bandwidth of the edge node according to the requirements of applications in different priority queues, aiming to reduce the queuing and processing delay of latency-sensitive streams while maximizing the edge-node throughput. At the network level, JANUS actively redirects incoming data streams to the less-loaded ones to achieve better network-wide load balance and improve the overall throughput. Experiments show that JANUS reduces the latency to only 16.6% of a non-priority based solution and improves the throughput by 1.7x of a state-of-the-art priority-aware data stream scheduling approach.

show abstract

Section: Prioritized Max Request First Served Heuristicmentioning

confidence: 99%

Section: Other Comparative Heuristicsmentioning

confidence: 99%