Methods for improving the availability of spot instances: A survey

Lin, Liduo; Pan, Li; Liu, Shijun

doi:10.1016/j.compind.2022.103718

Cited by 14 publications

(4 citation statements)

References 72 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, the uniform price cannot adapt to the varying supply-and-demand dynamics. On the contrary, AWS and Ali Cloud adopt the dynamic pricing strategy with an auction-like preemption rule 6,7 . In this strategy, consumers bid for resources, and preemption occurs when the real-time price exceeds their bids.…”

Section: A Pricing Strategies For the Excess Computing Capacitymentioning

confidence: 99%

“…Let J (π ) represents the expected total revenue in the whole horizon T . Therefore, the objective of the CNSMDP can be formulated as Equation (7).…”

Section: ) Maximum Preemption Rate Constraintmentioning

confidence: 99%

“…By solving for the Lagrange multiplier λ θ and corresponding pricing policy π θ in Equation ( 9), we can identify the optimal solution (λ * ,π * ) that satisfies the constraint conditions and maximizes the expected total revenue in the DPPCS problem stated in Equation (7). The optimal solution is denoted as…”

Section: Piecewise Lagrangian Dual Modelmentioning

confidence: 99%

“…This pricing strategy not only has the potential to maximize revenue but also reduces the probability of being preempted. In recent years, dynamic pricing has gained significant attention from cloud operators such as AWS and Ali Cloud [7]. Unfortunately, empirical evidence shows that the price fluctuations in dynamic pricing lack effective theoretical and methodological guidance, which leads to random price changes within a tight price range [8], [9].…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

A Novel Dynamic Pricing Approach for Preemptible Cloud Services

Peng,

Cheng

2023

IEEE Access

View full text Add to dashboard Cite

Dynamic pricing for preemptible cloud services (DPPCS) is highly demanded to effectively utilize the excess capacity in cloud computing. However, the dynamic nature of excess capacity exhibits high non-stationarity, which is characterized by multi-temporal stochastic patterns with time-varying statistical properties. The non-stationarity results in the DPPCS problem being a Non-Stationary Markov Decision Process (NSMDP) with unknown transition probabilities. Moreover, DPPCS is constrained by a certain maximum preemption rate, further complicating the DPPCS problem as a Constrained NSMDP (CNSMDP). We transform the CNSMDP into a piecewise Lagrangian dual model, which converts the CNSMDP into an unconstrained optimization problem. To solve the above problem, we propose a novel Q-Learning approach for DPPCS. We first present estimation methods for the unknown environment parameters, including a detection method for identifying temporal pattern changes, and a diffusion approximation method for estimating the actual preemption rate. Then, we introduce a Lagrange multiplier updating method, which can strike a balance between revenue and the preemption rate in the reward function. Building upon the above methods, we develop a Constrained Non-Stationary Q-Learning (CNSQL) algorithm for DPPCS, which dynamically adjusts its learning process to adapt to the multi-temporal patterns. Through simulated experiments, we demonstrate the effectiveness of our proposed approach compared to state-of-the-art algorithms. It performs well in improving revenue generated from excess capacity while maintaining the actual preemption rate within the specified constraint.

show abstract

Section: A Pricing Strategies For the Excess Computing Capacitymentioning

confidence: 99%

“…Let J (π ) represents the expected total revenue in the whole horizon T . Therefore, the objective of the CNSMDP can be formulated as Equation (7).…”

Section: ) Maximum Preemption Rate Constraintmentioning

confidence: 99%

Section: Piecewise Lagrangian Dual Modelmentioning

confidence: 99%