Beyond Poisson: Modeling Inter-Arrival Time of Requests in a Datacenter

Juan, Da-Cheng; Li, Lei; Peng, Hao‐Kai; Marculescu, Diana; Faloutsos, Christos

doi:10.1007/978-3-319-06605-9_17

Cited by 34 publications

(17 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It can be seen that it is hard to fit the inter-arrival times to a known distribution. A similar observation and a more detailed analysis on the job arrivals could be found in [17] as well. This further justifies the usefulness of our online algorithm, which works for arbitrary job arrivals.…”

Section: Trace Driven Simulation For Generic Concave Cost Job Schedulingsupporting

confidence: 78%

Online Resource Scheduling Under Concave Pricing for Cloud Computing

Zhang

et al. 2016

IEEE Trans. Parallel Distrib. Syst.

View full text Add to dashboard Cite

With the booming cloud computing industry, computational resources are readily and elastically available to the customers. In order to attract customers with various demands, most Infrastructure-as-a-service (IaaS) cloud service providers offer several pricing strategies such as pay as you go, pay less per unit when you use more (so called volume discount), and pay even less when you reserve. The diverse pricing schemes among different IaaS service providers or even in the same provider form a complex economic landscape that nurtures the market of cloud brokers. By strategically scheduling multiple customers' resource requests, a cloud broker can fully take advantage of the discounts offered by cloud service providers. In this paper, we focus on how a broker can help a group of customers to fully utilize the volume discount pricing strategy offered by cloud service providers through cost-efficient online resource scheduling. We present a randomized online stack-centric scheduling algorithm (ROSA) and theoretically prove the lower bound of its competitive ratio. Three special cases of the offline concave cost scheduling problem and the corresponding optimal algorithms are introduced. Our simulation shows that ROSA achieves a competitive ratio close to the theoretical lower bound under the special cases. Trace-driven simulation using Google cluster data demonstrates that ROSA is superior to the conventional online scheduling algorithms in terms of cost saving.

show abstract

Section: Trace Driven Simulation For Generic Concave Cost Job Schedulingsupporting

confidence: 78%

Online Resource Scheduling Under Concave Pricing for Cloud Computing

Zhang

et al. 2016

IEEE Trans. Parallel Distrib. Syst.

View full text Add to dashboard Cite

show abstract

“…We assume that the VMs IAT is exponentially distributed with mean of 1 minute, considering that we are dealing with data center of average access rates [24] [25]. The IAT spans from few seconds to reach maximum value of 5 minutes, and the serving durations are generated using the uniform distribution.…”

Section: Resultsmentioning

confidence: 99%

Energy efficient resource provisioning with VM migration heuristic for Disaggregated Server design

Ali

Al-Salim

Lawey

et al. 2016

2016 18th International Conference on Transparent Optical Networks (ICTON)

View full text Add to dashboard Cite

This article introduces an energy efficient heuristic that performs resource provisioning and Virtual Machine (VM) migration in the Disaggregated Server (DS) schema. The DS is a promising paradigm for future data centers where servers' components are disaggregated at the hardware unit levels and resources of similar type are combined in type respective pools, such as processing pools, memory pools and IO pools. We examined 1000 VM requests that demand various processing, memory and IO requirements. Requests have exponentially distributed inter arrival time and with uniformly distributed service duration periods. Resources occupied by a certain VM are released when the VM finishes its service duration. The heuristic optimises VMs allocation and dynamically migrates existing VMs to occupy newly released energy efficient resources. We assess the energy efficiency of the heuristic by applying increasing service duration periods. The results of the numerical simulation indicate that our power savings can reach up to 55% when compared to our pervious study where VM service duration is infinite and resources are not released. Keywords: Virtual Machine, VM Migration, Disaggregated Server, data center, energy consumption. INTRODUCTIONNowadays, virtualization within cloud computing are becoming prevalent technologies that greatly shape our lives, and accordingly, the energy consumption of the physical infrastructure that provides resources for the clouds is growing [1]. Thus, energy management is becoming a key challenge for data centers to reduce all their energy consumption and their total costs. A number of energy efficient data centre and inter data centre network architectures were proposed and studied in [2]- [7].Serious concerns are raised currently regarding traditional server and networking design and significant research efforts were dedicated to improve the design of switches and servers, leading to new paradigms for future data centers designs [8]- [19].In , and different disaggregated server architectures were introduced by presenting the memory and IO ports as separate blades from the server block allowing resources to be disaggregated across a system to enable data centers vertical elasticity. In [17] and [18] the energy sufficiency potential of the DS architecture was studied considering the resource provisioning and VM allocation. In [19] a comprehensive review of the DS was introduced.A DS based data centre, shown in Fig. 1, is a departure from the conventional (Single box Server) to disaggregated resources combined in resource pools. With DS, the different server resources are disaggregated from each other and combined with resources of the same type in resources pool constructing CPU pools, memory pools and IO pools. However the communication among resources within the same pool (Intra Rack Fabric in Fig. 1) or resources from different pools (Inter Rack Fabric in Fig. 1) is an important point that must be managed efficiently.In this paper, we consider a DS-based data center as a new data cente...

show abstract

“…[27] used periodicity and burstiness to specify and workloads in order to classify the workloads for elastic resource provision. [28] proposed a Hierarchical Bundling Model (HIBM) to model inter-arrival process. Our workload model falls into abstract mathematic workload modeling.…”

Section: Related Workmentioning

confidence: 99%

Resource Demand Forecasting Model Based on Dynamic Cloud Workload

C¹,

Zhou²

2017

Preprint

View full text Add to dashboard Cite

Abstract:The primary attraction of IaaS is providing elastic resources on demand. It becomes imperative that IaaS-users have an effective methodology for learning what resources they require, how many resources and for how long they need. However, the heterogeneity of resources, the diversity resource demands of different cloud applications and the variation of application-user behaviors pose IaaS-users big challenge. In this paper, we purpose a unified resource demand forecasting model suiting for different applications, various resources and diverse time-varying workload patterns. With the model, taking input from parameterized applications, resources and workload scenarios, the corresponding resources demands during any time interval can be deduced as output. The experiments configure concrete functions and parameters to help understanding the above model.

show abstract

Beyond Poisson: Modeling Inter-Arrival Time of Requests in a Datacenter

Cited by 34 publications

References 9 publications

Online Resource Scheduling Under Concave Pricing for Cloud Computing

Online Resource Scheduling Under Concave Pricing for Cloud Computing

Energy efficient resource provisioning with VM migration heuristic for Disaggregated Server design

Resource Demand Forecasting Model Based on Dynamic Cloud Workload

Contact Info

Product

Resources

About