Towards Faster Response Time Models for Vertical Elasticity

Lakew, Ewnetu Bayuh; Klein, Cristian; Hernández-Rodríguez, Francisco; Elmroth, Erik

doi:10.1109/ucc.2014.86

Cited by 37 publications

(21 citation statements)

References 13 publications

(11 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Vertical elasticity adds flexibility as it eliminates the overhead in starting a new VM and loading the service model. Prior efforts to scale the CPU resources vertically appear in [24], [25] including an approach that uses the discrete-time feedback controller leveraging MAPE-K loop for containerized applications [3]. Barista uses an efficient, proactive method to trigger the scaling of resources horizontally while relying on vertical scaling reactively to allocate and de-allocate CPU cores for model correction when our estimation model cannot predict accurately.…”

Section: Dynamic Infrastructure Elasticitymentioning

confidence: 99%

BARISTA: Efficient and Scalable Serverless Serving System for Deep Learning Prediction Services

Bhattacharjee

Chhokra

Kang

et al. 2019

2019 IEEE International Conference on Cloud Engineering (IC2E)

View full text Add to dashboard Cite

Pre-trained deep learning models are increasingly being used to offer a variety of compute-intensive predictive analytics services such as fitness tracking, speech and image recognition. The stateless and highly parallelizable nature of deep learning models makes them well-suited for serverless computing paradigm. However, making effective resource management decisions for these services is a hard problem due to the dynamic workloads and diverse set of available resource configurations that have various deployment and management costs. To address these challenges, we present a distributed and scalable deep-learning prediction serving system called Barista and make the following contributions. First, we present a fast and effective methodology for forecasting workloads by identifying various trends. Second, we formulate an optimization problem to minimize the total cost incurred while ensuring bounded prediction latency with reasonable accuracy. Third, we propose an efficient heuristic to identify suitable compute resource configurations. Fourth, we propose an intelligent agent to allocate and manage the compute resources by horizontal and vertical scaling to maintain the required prediction latency. Finally, using representative real-world workloads for an urban transportation service, we demonstrate and validate the capabilities of Barista.

show abstract

Section: Dynamic Infrastructure Elasticitymentioning

confidence: 99%

BARISTA: Efficient and Scalable Serverless Serving System for Deep Learning Prediction Services

Bhattacharjee

Chhokra

Kang

et al. 2019

2019 IEEE International Conference on Cloud Engineering (IC2E)

View full text Add to dashboard Cite

show abstract

“…As shown in Table 1, most previous horizontal scaling studies did not involve heterogeneous resources management, as horizontal scaling adds or removes fixed resource units. In contrast to horizontal scaling method, vertical scaling method [5,[20][21][22][23][24][25][26] manages resources by increasing or reducing the sizes of VMs.…”

Section: Related Workmentioning

confidence: 99%

“…However, this coarse-grained approach may lead to underutilization as a result of excessive allocation. Vertical scaling methods [5,[20][21][22][23][24][25][26] dynamically allocate resources by changing the resource level, such as the CPU and memory allocation for each VM. This fine-grained approach easily meets the specific resource demands for services.…”

Section: Introductionmentioning

confidence: 99%

An Adaptive Control Method for Resource Provisioning with Resource Utilization Constraints in Cloud Computing

Gong¹,

Yin²,

Zhang³

et al. 2019

IJCIS

View full text Add to dashboard Cite

A B S T R A C TCloud computing enables users to purchase virtual resources on demand; therefore, the service requests change over time. Dynamic resource provisioning for cloud computing has become a key challenge. To reduce the associated costs, resource utilization must be improved, and ensure the Quality of Service (QoS) to meet the service-level agreements (SLAs). However, it is difficult to improve the resource utilization level and maintain a high QoS with fluctuating workloads in shared cloud computing systems. In this paper, we propose an adaptive control method for resource provisioning in cloud computing systems to simultaneously improve the resource utilization, achieve a satisfactory QoS, and react to the dynamic workload. We proposed an approach integrating adaptive multi-input and multi-output (MIMO) control and radial basis function (RBF) neural network to react to the highly dynamic workloads, and precisely control the resource allocation to improve resource utilization based on the maximum allowable QoS requirements. Experiments based on real-world workloads show that the proposed approach jointly improves resource utilization, maintains a satisfactory QoS, and handles workload fluctuations in a coordinated manner.

show abstract

“…For the VM vertical elasticity, there are some works which focus on CPU resizing, e.g, (Lakew et al, 2014) and (Dawoud et al, 2012), while others concentrate on memory resizing, e.g., (Baruchi and Midorikawa, 2011) as well as combination of both such as the work of (Farokhi et al, 2015). (Monsalve et al, 2015) proposed an approach that controls CPU shares of a container, this approach uses CFS scheduling mode.…”

Section: Related Workmentioning

confidence: 99%

“…Vertical elasticity consists in increasing or decreasing characteristics of computing resources, such as CPU, memory, etc. (Lakew et al, 2014).…”

Section: Introductionmentioning

confidence: 99%

Coordinating Vertical Elasticity of both Containers and Virtual Machines

Al-Dhuraibi

Zalila

Djarallah³

et al. 2018

Proceedings of the 8th International Conference on Cloud Computing and Services Science

View full text Add to dashboard Cite

Abstract:Elasticity is a key feature in cloud computing as it enables the automatic and timely provisioning and deprovisioning of computing resources. To achieve elasticity, clouds rely on virtualization techniques including Virtual Machines (VMs) and containers. While many studies address the vertical elasticity of VMs and other few works handle vertical elasticity of containers, no work manages the coordination between these two vertical elasticities. In this paper, we present the first approach to coordinate vertical elasticity of both VMs and containers. We propose an auto-scaling technique that allows containerized applications to adjust their resources at both container and VM levels. This work has been evaluated and validated using the RUBiS benchmark application. The results show that our approach reacts quickly and improves application performance. Our coordinated elastic controller outperforms container vertical elasticity controller by 18.34% and VM vertical elasticity controller by 70%. It also outperforms container horizontal elasticity by 39.6%.

show abstract

Towards Faster Response Time Models for Vertical Elasticity

Cited by 37 publications

References 13 publications

BARISTA: Efficient and Scalable Serverless Serving System for Deep Learning Prediction Services

BARISTA: Efficient and Scalable Serverless Serving System for Deep Learning Prediction Services

An Adaptive Control Method for Resource Provisioning with Resource Utilization Constraints in Cloud Computing

Coordinating Vertical Elasticity of both Containers and Virtual Machines

Contact Info

Product

Resources

About