Stabilizing the virtual response time in single-server processor sharing queues with slowly time-varying arrival rates

Cho, Yongkyu; Ko, Young Myoung

doi:10.1007/s10479-019-03511-9

Cited by 4 publications

(4 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Wentao et al [27] studied optimal load balancing for a certain type of cloud architecture that well reflects machine learning applications. Kwon and Gautam [28] and Cho and Ko [29] investigated methods to time stabilize the performance of stochastic service systems that well-model cloud data centers. Anton et al [30] first showed that a redundancy system (e.g., MapReduce) could help improve the performance of data center computing in case the servers' capacities are sufficiently heterogeneous.…”

Section: B Real-time Operationmentioning

confidence: 99%

See 1 more Smart Citation

Power- and QoS-Aware Job Assignment With Dynamic Speed Scaling for Cloud Data Center Computing

Cho

2022

IEEE Access

Self Cite

View full text Add to dashboard Cite

As the usage of mission-critical mobile applications increases in Industry 4.0, such as smart manufacturing and self-driving cars, the cloud computing paradigm and its supporting data centers have become more crucial. However, a common practice in the cloud data center computing industry tends to supply a surfeit of computing resources mainly for a robust quality-of-service (QoS). In this paper, we propose a simple real-time algorithm which combines a power-aware job assignment policy for a centralized job dispatcher and a power-and QoS-aware dynamic speed scaling policy for each physical machine (PM). The job assignment policy is called "Join the Least Power Consuming (LPC) Server" that routes an incoming cloud job to a server spending minimum power upon request. The server-side adaptive speed scaling policy expedites energy efficiency and satisfies response time-associated QoS condition. We call this policy "Minimizing Earliness (ME)" since it manages the server speed towards finishing jobs at their deadlines as precisely as possible, reducing the earliness of job completion. The design principle of LPC-ME combination supports both energy efficiency and service quality required in cloud data centers. Numerical experiments compare the proposed algorithm's power consumption and response time with those of existing popular policies and demonstrate better energy efficiency with negligible degradation of service quality.INDEX TERMS Cloud computing, data centers, energy efficiency, queueing analysis, resource management.

show abstract

Section: B Real-time Operationmentioning

confidence: 99%

“…Attaching the subscript i, the applicationspecific workload is now expressed by w i (t) = β i λ i (t). See Cho and Ko [29] for more details about the NSNP and GI t /GI t /1/P S model.…”

Section: ) Workload Related Parametersmentioning

confidence: 99%

Power- and QoS-Aware Job Assignment With Dynamic Speed Scaling for Cloud Data Center Computing

Cho

2022

IEEE Access

Self Cite

View full text Add to dashboard Cite

show abstract

“…Tirdad et. al [16] Cho and Ko [17] investigate the stabilization of the mean virtual response time in a single-server processor sharing (PS) queueing system with a time-varying arrival rate and a service rate control. The authors propose and compare a modified square-root (SR) control and a difference-matching (DM) control to stabilize the mean virtual response time.…”

Section: Nonstationarity Of Arrivals In Call Centersmentioning

confidence: 99%

Nonstationarity and Abandonment in Markovian Queues with Application to Call Centers

Mahajan¹

2020

AJOMIS

View full text Add to dashboard Cite

Both large and small firms maintain call centers to establish contact between themselves and their customers. A call center is staffed by Customer Service Representatives (CSRs). In addition to CSRs, a call center needs computers and telecommunication equipment such as an Automatic Call Distributor (ACD). If calls arrive according to a Poisson arrival process and the service times have an exponential distribution, the call center can be modeled as an M/M/s queue where s is the number of CSRs. Typical calculations include finding the number of CSRs required and finding the number of trunks lines required. However, if call center models ignore the abandonment behavior of customers, they distort information that is relevant to management. Typically, ignoring abandonment would lead to overstaffing, as fewer CSRs are actually needed, because some of the callers abandon the system. Also, nonstationarity of arrivals is highly prevalent in call centers. Green, Kolesar and Whitt plot hourly arrival rates for a financial services call center. There is significant variation in arrivals by time of day. In this paper, we model call centers as multiserver Markovian queues with both nonstationarity and abandonment. Nonstationarity is modeled by having a Poisson arrival stream with time dependent mean, which varies according to a sinusoid. Abandonment is defined in terms of an exponential patience random variable, which is the amount of time the caller would be on hold before abandoning the call. We numerically study the performance of this nonstationary M(t)/M/s+M queue with abandonment and compare its performance measures with those of the stationary M/M/s+M queue. We find that approximating a nonstationary system by a stationary system, even at low levels of nonstationarity, can lead to significant errors. Similar results in a system without abandonment, have been obtained by Green, Kolesar and Svoronos. Additionally, we find that abandonment dampens the effect of nonstationarity. Since abandonment and nonstationarity are both present together in real call centers, a real call center with a high level of abandonment behaves closer to an ideal stationary system.

show abstract

“…Cho and Ko investigated the stabilization of the mean virtual response time in a single‐server processor‐sharing (PS) queueing system with a time‐varying arrival rate and a service rate control (GIt/GIt/1/PS queue). They proposed a mechanism to stabilize the mean virtual response time of a GIt/GIt/1/PS queue.…”

Section: Primary Studies and Classification Schemesmentioning

confidence: 99%

Applying queue theory for modeling of cloud computing: A systematic review

Ghomi

Rahmani

Qader

2019

Concurrency and Computation

View full text Add to dashboard Cite

The cloud computing paradigm is an important service in the Internet for sharing and providing resources in a cost-efficient way. Modeling of a cloud system is not an easy task because of the complexity and large scale of such systems. Cloud reliability could be improved by modeling the various aspects of cloud systems, including scheduling, service time, wait time, and hardware and software failures. The aim of this study is to survey research studies done on the modeling of cloud computing using the queuing system in order to identify where more emphasis should be placed in both current and future research directions. This paper follows the goal by investigating the articles published between 2008 and January 2017 in journals and conferences. A systematic mapping study combined with a systematic literature review was performed to find the related literature, and 71 articles were selected as primary studies that were classified in relation to the focus, research type, and contribution type. We classified the modeling techniques of cloud computing using the queuing theory in seven categories based on their focus area: (1) performance, (2) quality of service, (3) workflow scheduling, (4) energy savings, (5) resource management, (6) priority-based servicing, and (7) reliability. A majority of the primary articles focus on performance (37%), 15% of them focus on resource management, 14% of them focus on quality of service, 13% of them focus on workflow scheduling, 13% of them focus on energy savings, 4% of them focus on priority-based servicing for requests, and 4% of them focus on reliability. This work summarizes and classifies the research efforts conducted on applying queue theory for modeling of cloud computing (AQTMCC), providing a good starting point for further research in this area. KEYWORDScloud computing, cloud modeling, queuing systems, queuing theory, systematic review INTRODUCTIONCloud computing is the recent evolution in the information technology and is growing very fast due to the extensive use of mobile devices, such as PDAs, cell phones, and tablets. 1 To emphasize the goals of cloud computing, we refer to the definition of NIST (National Institute of Standards and Technology); a cloud model is composed of five characteristics, three service models, and four deployment models. 2 Service models of cloud computing are as follows: (1) Software as a Service (SaaS), such as CRM (Customer Relationship Management), virtual desktop, communications, and games; (2) Platform as a Service (PaaS), such as databases, web servers, and deployment tools; and (3) Infrastructure as a Service (IaaS), such as virtual machines (VMs), storages, servers, and load balancers. Deployment models include private clouds, public clouds, community clouds, and hybrid clouds. The characteristics of a cloud are on-demand self-servicing, elasticity, broad network access, resource pooling, and measured service. 1 A data center in cloud computing consists of several servers that are organized in racks and connected through communicati...

show abstract

Stabilizing the virtual response time in single-server processor sharing queues with slowly time-varying arrival rates

Cited by 4 publications

References 26 publications

Power- and QoS-Aware Job Assignment With Dynamic Speed Scaling for Cloud Data Center Computing

Power- and QoS-Aware Job Assignment With Dynamic Speed Scaling for Cloud Data Center Computing

Nonstationarity and Abandonment in Markovian Queues with Application to Call Centers

Applying queue theory for modeling of cloud computing: A systematic review

Contact Info

Product

Resources

About