DRAGON: Decentralized Fault Tolerance in Edge Federations

Tuli, Shreshth; Casale, Giuliano; Jennings, Nicholas R.

doi:10.1109/tnsm.2022.3199886

Cited by 10 publications

(5 citation statements)

References 56 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This innovation has subsequently attracted widespread scholarly attention. For instance, an efficient deep learning model based on memory was proposed to address the challenges of deploying latency-critical and AI-based resource-intensive applications within the constrained devices of edge federation systems [2]. Additionally, a scheduling strategy that allocates incoming tasks based on energy consumption levels was developed to improve load balancing in terms of energy consumption and heat dissipation within the edge federation environment [3].…”

Section: Related Workmentioning

confidence: 99%

Tripartite Game Theory-Based Edge Resource Pricing Approach for Edge Federation

Ye,

Feng,

2024

Preprint

View full text Add to dashboard Cite

Mobile Edge Computing (MEC) deploys edge servers close to end-users to address their computing requirements, with an emphasis on minimizing latency and enhancing security. However, the significant deployment costs associated with edge servers and the potential for resource wastage due to redundant deployment by various Edge Infrastructure Providers (EIPs) highlight the importance of adopting edge federation architecture. This architectural approach aims to deliver superior computing services tailored for MEC scenarios. Nevertheless, effective resource allocation and pricing strategies are crucial to ensure optimal resource sharing within the edge federation. This paper presents a game-theoretic model involving three primary stakeholders: terminal users, resource demanders within the federation, and resource providers within the federation. The game among these stakeholders is structured into two levels, with negotiation rounds playing a pivotal role. At the first level, the game outlines the initial resource demands of terminal users and the initial pricing by EIPs, while also considering the influence of historical pricing trends on resource demands. The second level focuses on defining the psychological expectations of resource demanders within the federation and the initial pricing by resource providers. To solve this game model, this paper introduces the Tripartite Game Pricing Approach (TGPA). Simulation results corroborate that TGPA outperforms existing pricing approaches, yielding higher total revenue for the edge federation and ensuring a certain utilization of computational resources within acceptable time constraints.

show abstract

Section: Related Workmentioning

confidence: 99%

Tripartite Game Theory-Based Edge Resource Pricing Approach for Edge Federation

Ye,

Feng,

2024

Preprint

View full text Add to dashboard Cite

show abstract

“…Aral et al [6] proposed a machine learning method to predict the failure probability of Edge servers by minimizing service failure, redundancy cost, and network latency. Tuli et al [7] proposed a decentralized fault tolerance method for Edge service to reduce SLA violations. c) Limitations: These related works assess devices based on their SLA violation only, ignore service success, and give equal weights to the different QoS metrics.…”

Section: Related Workmentioning

confidence: 99%

“…An incorrect resource selection in such a complex and hardly predictable environment leads to service interruptions, performance degradation, and profit reduction. The demand for service level agreements (SLA) becomes increasingly important to prevent QoS violations [3]- [7] and gives performance indicators for future resource selection.…”

Section: Introductionmentioning

confidence: 99%

“…Existing works [8]- [12] focus primarily on SLA management in the Cloud, and only a few recent efforts [13]- [19] perform a quantitative resource analysis based on historical data [3]- [7] to predict violations on the computing continuum. To address this need, we propose a proactive SLAaware (PROS) method relying on distributed monitoring to reduce SLA violations and increase service success on the computing continuum.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Proactive SLA-aware Application Placement in the Computing Continuum

Samani¹,

Mehran²,

Kimovski³

et al. 2023

2023 IEEE International Parallel and Distributed Processing Symposium (IPDPS)

View full text Add to dashboard Cite

The accelerating growth of modern distributed applications with low delivery deadlines leads to a paradigm shift towards the multi-tier computing continuum. However, the geographical dispersion, heterogeneity, and availability of the continuum resources may result in failures and quality of service degradation, significantly negating its advantages and lowering users' satisfaction. We propose in this paper a proactive application placement (PROS) method relying on distributed coordination to prevent the quality of service violations through service-level agreements on the computing continuum. PROS employs a sigmoid function with adaptive weights for the different parameters to predict the service level agreement assurance of devices based on their past credentials and current capabilities. We evaluate PROS using two application workloads with different traffic stress levels up to 90 million services on a real testbed with 600 heterogeneous instances deployed over eight geographical locations. The results show that PROS increases the success rate by 7 %-33 %, reduces the response time by 16 %-38 %, and increases the deadline satisfaction rate by 19 %-42 % compared to two related work methods. A comprehensive simulation study with 1000 devices and a workload of up to 670 million services confirm the scalability of the results.

show abstract

“…Further, γ is a weighting parameter that can be set by a user based on the deployment scenario and as per the relative importance of the two metrics for the user. Weighting schemes are common to reduce multi-objective optimization to more efficiently solvable single-objective problems [47], [48], [49]. In particular a convex combination allows us to efficiently combine both metrics.…”

Section: E Formulationmentioning

confidence: 99%

MetaNet: Automated Dynamic Selection of Scheduling Policies in Cloud Environments

Tuli

Casale

Jennings

2022

2022 IEEE 15th International Conference on Cloud Computing (CLOUD)

View full text Add to dashboard Cite

Intelligent Virtual Machine (VM) provisioning is central to cost and resource efficient computation in cloud computing environments. As bootstrapping VMs is time-consuming, a key challenge for latency-critical tasks is to predict future workload demands to provision VMs proactively. However, existing AI-based solutions tend to not holistically consider all crucial aspects such as provisioning overheads, heterogeneous VM costs and Quality of Service (QoS) of the cloud system. To address this, we propose a novel method, called CILP, that formulates the VM provisioning problem as two sub-problems of prediction and optimization, where the provisioning plan is optimized based on predicted workload demands. CILP leverages a neural network as a surrogate model to predict future workload demands with a co-simulated digital-twin of the infrastructure to compute QoS scores. We extend the neural network to also act as an imitation learner that dynamically decides the optimal VM provisioning plan. A transformer based neural model reduces training and inference overheads while our novel two-phase decision making loop facilitates in making informed provisioning decisions. Crucially, we address limitations of prior work by including resource utilization, deployment costs and provisioning overheads to inform the provisioning decisions in our imitation learning framework. Experiments with three public benchmarks demonstrate that CILP gives up to 22% higher resource utilization, 14% higher QoS scores and 44% lower execution costs compared to the current online and offline optimization based state-of-the-art methods.

show abstract

DRAGON: Decentralized Fault Tolerance in Edge Federations

Cited by 10 publications

References 56 publications

Tripartite Game Theory-Based Edge Resource Pricing Approach for Edge Federation

Tripartite Game Theory-Based Edge Resource Pricing Approach for Edge Federation

Proactive SLA-aware Application Placement in the Computing Continuum

MetaNet: Automated Dynamic Selection of Scheduling Policies in Cloud Environments

Contact Info

Product

Resources

About