An Architecture for Automatic Scaling of Replicated Services

Aniello, Leonardo; Bonomi, Silvia; Lombardi, Federico; Zelli, Alessandro; Baldoni, Roberto

doi:10.1007/978-3-319-09581-3_9

Cited by 15 publications

(13 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Service time is the time a server spends on processing the request, which is widely used in the queuing models to approximate the average response time or sojourn time. Except for a few works [Gergin et al 2014;Han et al 2014] that assume this metric as known a priori, to accurately measure it, either offline profiling [Prodan and Nae 2009] or support from the application [Aniello et al 2014] is required. Therefore, instead of directly probing it, some works use other approaches to approximate it.…”

Section: High-level Metricsmentioning

confidence: 99%

“…[Kamra et al 2004], [Villela et al 2007], [Sharma et al 2012], [Gandhi et al 2014a;Gandhi et al 2014b], and [Gergin et al 2014] employed this method. Some described the cluster using a queue with multiple servers, like [Ali-Eldin et al 2012b], [Jiang et al 2013], [Aniello et al 2014], and [Han et al 2014]. Other works modeled each server as a separate queue, such as [Doyle et al 2003], [Urgaonkar et al 2008], [Roy et al 2011], [Ghanbari et al 2012], [Kaur and Chana 2014], [Spinner et al 2014], and [Jiang et al 2010].…”

Section: Analytical Modelingmentioning

confidence: 99%

See 1 more Smart Citation

Auto-Scaling Web Applications in Clouds

2018

View full text Add to dashboard Cite

Web application providers have been migrating their applications to cloud data centers, attracted by the emerging cloud computing paradigm. One of the appealing features of the cloud is elasticity. It allows cloud users to acquire or release computing resources on-demand, which enables web application providers to automatically scale the resources provisioned to their applications without human intervention under a dynamic workload to minimize resource cost while satisfying Quality of Service (QoS) requirements. In this paper, we comprehensively analyze the challenges that remain in auto-scaling web applications in clouds and review the developments in this field. We present a taxonomy of auto-scalers according to the identified challenges and key properties. We analyze the surveyed works and map them to the taxonomy to identify the weaknesses in this field. Moreover, based on the analysis, we propose new future directions that can be explored in this area.

show abstract

Section: High-level Metricsmentioning

confidence: 99%

Section: Analytical Modelingmentioning

confidence: 99%

Auto-Scaling Web Applications in Clouds

2018

View full text Add to dashboard Cite

show abstract

“…This article describes a cloud elasticity initiative called AutoElastic, which is evaluated using the proposed metrics. Thus, we organized this section in two groups: (i) performance and/or energy‐driven cloud computing middleware and models ; (ii) metrics to measure energy consumption and cost in cloud environments . These groups are addressed in Sections 2.1 and 2.2.…”

Section: Related Workmentioning

confidence: 99%

“…By combining feed‐forward control and feedback control, ElastMan addresses the challenges of the variable performance of cloud VMs. Aniello et al developed an architecture for automatically scaling replicated services. A queuing model of the replicated service is used to compute the expected response time, given the current configuration (number of replicas) and the distributions of both input requests and service times.…”

Section: Related Workmentioning

confidence: 99%

“…Although elasticity allows applications to acquire and release resources dynamically, adjusting to changing demands, deciding about appropriate thresholds and measuring performance and energy consumption accurately are not easy tasks . Our analysis shows that related work presents a gap concerning defining decision functions for elasticity viability in the combination of performance and energy, besides not addressing a discussion of lower and upper thresholds . Responding to this motivation, we developed an elasticity model called AutoElastic, which automatically reorganizes resources for loop‐based synchronous parallel applications.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Joint‐analysis of performance and energy consumption when enabling cloud elasticity for synchronous HPC applications

Righi

Costa

Rodrigues

et al. 2015

Concurrency and Computation

View full text Add to dashboard Cite

Summary A key characteristic of cloud computing is elasticity, automatically adjusting system resources to an application's workload. Both reactive and horizontal approaches represent traditional means to offer this capability, in which rule‐condition‐action statements and upper and lower thresholds occur to instantiate or consolidate compute nodes and virtual machines. Although elasticity can be beneficial for many HPC (high‐performance computing) scenarios, it also imposes significant challenges in the development of applications. In addition to issues related to how we can incorporate this new feature in such applications, there is a problem associated with the performance and resource pair and, consequently, with energy consumption. Further exploring this last difficulty, we must be capable of analyzing elasticity effectiveness as a function of employed thresholds with clear metrics to compare elastic and non‐elastic executions properly. In this context, this article explores elasticity metrics in two ways: (i) the use of a cost function that combines application time with different energy models; (ii) the extension of speedup and efficiency metrics, commonly used to evaluate parallel systems, to cover cloud elasticity. To accomplish (i) and (ii), we developed an elasticity model known as AutoElastic, which reorganizes resources automatically across synchronous parallel applications. The results, obtained with the AutoElastic prototype using the OpenNebula middleware, are encouraging. Considering a CPU‐bound application, an upper threshold close to 70% was the best option for obtaining good performance with a non‐prohibitive elasticity cost. In addition, the value of 90% for this threshold was the best option when we plan an efficiency‐driven execution. Copyright © 2015 John Wiley & Sons, Ltd.

show abstract

Towards a Non-intrusive Recognition of Anomalous System Behavior in Data Centers

Baldoni

Cerocchi²,

Ciccotelli

et al. 2014

Lecture Notes in Computer Science

Self Cite

View full text Add to dashboard Cite

In this paper we propose a monitoring system of a data center that is able to infer when the data center is getting into an anomalous behavior by analyzing the power consumption at each server and the data center network traffic. The monitoring system is non-intrusive in the sense that there is no need to install software on the data center servers. The monitoring architecture embeds two Elman Recurrent Networks (RNNs) to predict power consumed by each data center component starting from data center network traffic and viceversa. Results obtained along six mounts of experiments, within a data center, show that the architecture is able to classify anomalous system behaviors and normal ones by analyzing the error between the actual values of power consumption and network traffic and the ones inferred by the two RNNs.

show abstract

An Architecture for Automatic Scaling of Replicated Services

Cited by 15 publications

References 30 publications

Auto-Scaling Web Applications in Clouds

Auto-Scaling Web Applications in Clouds

Joint‐analysis of performance and energy consumption when enabling cloud elasticity for synchronous HPC applications

Towards a Non-intrusive Recognition of Anomalous System Behavior in Data Centers

Contact Info

Product

Resources

About