Performance of Communication- and Computation-Intensive SaaS on the OpenStack Cloud

Bystrov, Oleg; Pacevič, Ruslan; Kačeniauskas, Arnas

doi:10.3390/app11167379

Cited by 12 publications

(15 citation statements)

References 45 publications

(56 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As shown in Figure 1, Mobile Edge Computing is located at the side of the roadside unit. At the same time, the MEC server has much more computing resources than the vehicle local resources, which can provide a good solution to the time-delay sensitive problem of computation-intensive tasks [5].…”

Section: Introductionmentioning

confidence: 99%

Edge Computing Resource Allocation and Optimization Method and Its Application in Internet of Vehicles Environment

Geng

2022

Security and Communication Networks

View full text Add to dashboard Cite

With the emergence of new vehicle-mounted services, such as autonomous driving and augmented reality, the computing capability at the edge of the Internet of vehicles (IoV) has become increasingly demanding, and the edge computing of the IoV has also emerged. With the rise of the IoV, new vehicle-mounted applications keep emerging. Vehicle-mounted applications need to process the sensor data of themselves and other vehicles around them. The amount of data is huge, and processing these sensor data requires powerful computing power. But typical vehicles have limited computing power, so such computation-intensive tasks need to be uploaded to a data center for processing. The MEC (Mobile Edge Computing) technology is a good choice to solve this problem. This paper advocates reasonable scheduling of parked vehicles to cooperate with edge servers for collaborative edge computing and studies the resource scheduling scheme under the target of user relative satisfaction. This paper describes the main entities and related functions in the network and proposes a secure and reliable protocol for interaction between entities. The serviceability of different parked vehicles in collaborative edge computing was evaluated to select stable task performers and solve the resource scheduling optimization problem between parked vehicles and edge servers. The simulation results show that the proposed scheme has better customer satisfaction and maximum overhead than the existing scheme that only considers using edge server to perform tasks. Therefore, the method of this paper has important theoretical significance and potential practical value.

show abstract

Section: Introductionmentioning

confidence: 99%

Edge Computing Resource Allocation and Optimization Method and Its Application in Internet of Vehicles Environment

Geng

2022

Security and Communication Networks

View full text Add to dashboard Cite

show abstract

“…The measured speedup values are close to those attained for relevant numbers of nodes on high latency 1 Gbps Ethernet networks as in parallel performance studies of DEM software. 9,15 Figure 2B presents weak scalability of the parallel software, when the increasingly larger problems are solved proportionally increasing the number of containers. Thus, benchmarks of 256,000, 500,000, 1,000,188, and 1,997,199 particles are solved on configurations of containers 1 × 4, 2 × 4, 4 × 4, and (5 + 3) × 4, respectively.…”

Section: Performance Predictionmentioning

confidence: 99%

“…benchmark for performance measurements. 9,15 The solution domain is assumed to be a cubic container with the 1.0 m-long edges. Half of the domain is filled with monosized particles, using a cubic structure.…”

Section: (A) (B)mentioning

confidence: 99%

“…Open‐source DEM code MercuryDPM 14 also has been deployed on cloud computing platform. Bystrov et al 15 have developed parallel MPI‐based DEM software for distributed‐memory architectures and deployed SaaS on the OpenStack cloud. However, the parallel SaaS might suffer from severe performance degradation due to higher latencies of networks, irregular domain partitions, substantial load imbalance, heterogeneity of virtual resources and inappropriately selected configurations of resources 1 .…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Cost‐ and performance‐aware resource selection for parallel software on heterogeneous cloud

Bystrov

Pacevič

Kačeniauskas

2023

Concurrency and Computation

Self Cite

View full text Add to dashboard Cite

SummaryCloud providers offer flexible infrastructures and on‐demand services, including the capability to deploy low cost virtual resources of many different types. However, the diversity of cloud resources followed by the important trade‐off between cost and performance makes the resource selection a challenging task for users in the case of parallel communication‐intensive software. The paper presents cost‐ and performance‐aware resource selection for parallel discrete element method (DEM) software as a service (SaaS) on heterogeneous OpenStack cloud. The developed resource selection uses preliminary application‐specific benchmarks of size smaller than targeted problems and the performance prediction based on speedup of parallel computations to obtain Pareto optimal solutions and to select the best configuration of containers from user's perspective. Hybrid parallelization of DEM software is developed by using OpenCL for shared‐memory multi‐core architectures and MPI for internode communications on distributed‐memory computer clusters. Round up and proportional pricing schemes are examined and compared from a user's perspective. Lower cost of computations obtained by using the proportional pricing scheme is always preferable for users. However, the difference approaches 1.0% of the cost calculated by using proportional pricing scheme, when long lasting computations are performed. The prediction tends to underestimate the execution time of DEM SaaS, but its accuracy is sufficient to obtain the same Pareto optimal solutions by using measured and predicted execution times. Pareto front and linear scalarization propose to select configurations of containers capable of exploiting higher memory bandwidth, which is specific to memory bandwidth bound DEM computations.

show abstract

“…The article [8] compares the features of Docker virtualization to those of server virtualization, but does not provide a performance comparison. Articles [9], [10], [11] compare the performance of Docker and KVM systems. [12] compares the performance of Docker and LXD systems.…”

Section: Introductionmentioning

confidence: 99%

Stress testing of Docker containers running on a Windows operating system

Sergeev

Rezedinova

Khakhina

2022

J. Phys.: Conf. Ser.

View full text Add to dashboard Cite

At the moment, Docker technology is becoming more prevalent in the Windows environment. The pertinent topic in this regard is how Docker containers running on the Windows operating system would behave in the event of a resource deficit. The article discusses circumstances in which there is insufficient processor or RAM to service all running containers. It is proposed to employ stress and volume testing to investigate the stability and reliability of Docker containers. The purpose of stress testing is to determine how the system behaves when the load on the application is greatly increased in comparison to what was initially expected. By delivering enormous amounts of data for processing, volume testing is accomplished. To assess container performance during stress testing, a Prometheus-based technology stack is used. Docker containers running in a Windows system have been shown to operate predictably under extreme load conditions. As the processor load increases, the performance of the containerized program drops proportionately to the decrease in processor time allocated. When there is insufficient RAM to run all containers, Docker uses virtual memory to dynamically distribute memory amongst containers. If a container's execution is halted due to a lack of available RAM on the system, the container's execution is terminated. Other containers continue to function normally.

show abstract

Performance of Communication- and Computation-Intensive SaaS on the OpenStack Cloud

Cited by 12 publications

References 45 publications

Edge Computing Resource Allocation and Optimization Method and Its Application in Internet of Vehicles Environment

Edge Computing Resource Allocation and Optimization Method and Its Application in Internet of Vehicles Environment

Cost‐ and performance‐aware resource selection for parallel software on heterogeneous cloud

Stress testing of Docker containers running on a Windows operating system

Contact Info

Product

Resources

About