Performance comparison between hypervisor- and container-based virtualizations for cloud users

Wang, Bo; Song, Ying; Cui, Xiao; Cao, Jie

doi:10.1109/icsai.2017.8248375

Cited by 11 publications

(5 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Each measurement procedure was run multiple times to account system uncertainty [24], in this case we run it for 30 times. To calculate performance overhead, we use overhead ratio [25] as defined in (1).…”

Section: Resultsmentioning

confidence: 99%

See 1 more Smart Citation

Performance analysis of multi services on container Docker, LXC, and LXD

2020

View full text Add to dashboard Cite

The emergence of the container in various cloud platforms from Open Stack to Google Cloud Platform has marked the industry interest in opting for container as their cloud service solution. However, the cloud users should aware of performance overheads of different virtualization solutions in order to avoid quality of service degradation because different container platforms delivered different performances. This research evaluated how different container platforms (Docker, LXC, and LXD) impacted in running different TCP services and also measured system performance of each container compared to the native system without any container solution based on overall performance metrics. This research focuses on the three most used PaaS: FTP Server, Web Server, and Mail Server. Related to previous works, our evaluation results show that performance could vary between containers. In terms of system performance, LXD shows better performance while server performance result varies depending on what service is being evaluated.

show abstract

Section: Resultsmentioning

confidence: 99%

“…Virtualization has played a great role in the growth of cloud computing and considered as one of foundation technology for clouds [1]. There are two kinds of virtualization solution, hypervisor and container.…”

Section: Introductionmentioning

confidence: 99%

Performance analysis of multi services on container Docker, LXC, and LXD

2020

View full text Add to dashboard Cite

show abstract

“…Penelitian yang dilakukan oleh Li, Kihl, Lu, & Andersson, (2017), juga menunjukkan ratarata performa yang dihasilkan oleh container secara umum lebih baik dari pada hypervisor hingga dapat dibandingkan langsung dengan mesin fisiknya. Namun berdasarkan penelitian yang dilakukan oleh Raho, Spyridakis, Paolino, & Raho, (2015) dan Wang, Song, Cui, & Cao, (2017) menunjukkan bawa perbedaan performa antara kedua jenis virtualisasi bisa bervariasi tergantung dari jenis pekerjaan yang ditugaskan kepada masing-masing virtualisasi dan tidak menemukan perbedaan performa yang besar antara kedua virtualisasi.…”

Section: Pendahuluanunclassified

Analisis Perbandingan Performa Virtualisasi Berbasis Container Dengan Virtualisasi Berbasis Hypervisor

Prasetyo

Benny

2021

SCIENCE TECH

View full text Add to dashboard Cite

Naskah diterima:08/01/2021Usage of virtualization for company is nothing new in this era. Virtualization helps company in reducing the resource they needed for creating new different servers. Among the virtualization technology that exists, there are two virtualization technologies that dominated the market, Hypervisor and container. Hypervisor is a virtualization technology which allow machine to create different virtual machines with different operating system, and container is a virtualization technology which isolated in operating system level. With different architecture, the performances between the two virtualization are considered different too. Two identical machines are created in cloud with performance test ran on both machines using apache jmeter to find out the performance between the two virtualization technology. The key variable in this performance test are response time, deviation, throughput, error, and CPU Utilization, with two different setting in each test, the first with 200 users and the second with 500 users. In the load of 200 users, machine using Hypervisor virtualization technology performs the same as the machine using container virtualization technology. However at load of 500 users, machine using Hypervisor virtualization technology fails to keep up and lose a significant amount of performances than the machine using container virtualization technology.

show abstract

“…One of the main problems with SIP network is the overload incurred by SIP proxy. Paper [9] presented a comparative study between hypervisor and docker. The hypervisor was chosen as Xen; further the overhead involved in virtualization in HPC and OLTP were discussed.…”

Section: Related Workmentioning

confidence: 99%

Segregating Signaling and Media Planes into Different containers

V¹,

N²

2021

JUSST

View full text Add to dashboard Cite

Scalability is an important aspect of communication networks. With the ascent of SIP and associated modern real-time protocols, IP telephony has become a revolutionary technology in connecting users through real-time voice communications with enabled video and instant messaging capabilities. B2BUA is a SIP server that provides call management and authentication functionality by reformulating the request and routing the traffic to other user agents in the network. It comprises signaling and media entities that handle all control signaling messages and real-time data(media) information respectively. The signaling and entities run as different processes in the same container. Such an architecture encounters a large CPU utilization after a specific number of maximum calls due to increase traffic flowing within the same node. Further Packet processing is CPU intensive and there is a need for an architecture that scales well with increasing traffic without hitting the CPU performance. The paper presents the design of decoupled architecture for Signaling and Media entities by running both the processes in different containers. With such an approach, one Signaling entity can communicate with multiple Media entities or vice versa thereby providing a suitable scalable solution to deal with the increased traffic and further maintaining the system efficiency. The paper is concluded by highlighting the difference between Kubernetes and OpenStack for the proposed architecture

show abstract

Performance comparison between hypervisor- and container-based virtualizations for cloud users

Cited by 11 publications

References 20 publications

Performance analysis of multi services on container Docker, LXC, and LXD

Performance analysis of multi services on container Docker, LXC, and LXD

Analisis Perbandingan Performa Virtualisasi Berbasis Container Dengan Virtualisasi Berbasis Hypervisor

Segregating Signaling and Media Planes into Different containers

Contact Info

Product

Resources

About