SDN-based server clusters with dynamic load balancing and performance improvement

Chiang, Mei‐Ling; Cheng, Hui-Sheng; Liu, Hsien-Yi; Chiang, Ching-Yi

doi:10.1007/s10586-020-03135-w

Cited by 28 publications

(17 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Currently, the VMC/H system is constructed using LVS cluster technology, KVM virtualization technology, HDFS, and Hbase distributed non-relational database. Our future work will compare its performance with that of a VMC/H system using software-defined networking [16][17][18] to design virtual server clusters. We will also study the use of other database technologies for sensor data storage, such as Cassandra [33], to increase the performance of the overall system.…”

Section: Discussionmentioning

confidence: 99%

“…Several studies [16][17][18] design their VM server clusters using software-defined networking (SDN). As load balancing is also critical for SDN-based server clusters, some propose new load balancing strategies to utilize cluster resources effectively.…”

Section: Related Workmentioning

confidence: 99%

“…As load balancing is also critical for SDN-based server clusters, some propose new load balancing strategies to utilize cluster resources effectively. Among them, the dynamic weighted random selection (DWRS) [18] considers the real-time server loads when dispatching requests to servers with different processing capabilities. Server loads are periodically collected and reported to the controller, and server weights are dynamically updated.…”

Section: Related Workmentioning

confidence: 99%

See 2 more Smart Citations

A Scalable Virtualized Server Cluster Providing Sensor Data Storage and Web Services

Chiang

Hou

2020

Symmetry

Self Cite

View full text Add to dashboard Cite

With the rapid development of the Internet of Things (IoT) technology, diversified applications deploy extensive sensors to monitor objects, such as PM2.5 air quality monitoring. The sensors transmit data to the server periodically and continuously. However, a single server cannot provide efficient services for the ever-growing IoT devices and the data they generate. This study bases on the concept of symmetry of architecture and quantities in system design and explores the load balancing issue to improve performance. This study uses the Linux Virtual Server (LVS) and virtualization technology to deploy a virtual machine (VM) cluster. It consists of a front-end server, also a load balancer, to dispatch requests, and several back-end servers to provide services. These receive data from sensors and provide Web services for browsing real-time sensor data. The Hadoop Distributed File System (HDFS) and HBase are used to store the massive amount of received sensor data. Because load-balancing is critical for resource utilization, this study also proposes a new load distribution algorithm for VM-based server clusters that simultaneously provide multiple services, such as sensor services and Web service. It considers the aggregate load of all back-end servers on the same physical server that provides multiple services. It also considers the difference between physical machines and VMs. Algorithms such as those for LVS, which do not consider these factors, can cause load imbalance between physical servers. The experimental results demonstrate that the proposed system is fault tolerant, highly scalable, and offers high availability and high performance.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

A Scalable Virtualized Server Cluster Providing Sensor Data Storage and Web Services

Chiang

Hou

2020

Symmetry

Self Cite

View full text Add to dashboard Cite

show abstract

“…Because traffic is unpredictable, networks such as data centers are dynamic, and traffic congestion can occur at any time. Chiang et al (2021) offered an SDN-based server cluster with a dynamic load balancing scheme that can be applied to a variety of data center network topologies. During flow transmissions, the recommended approach enables path changes.…”

Section: Introductionmentioning

confidence: 99%

Convergence time aware switch migration algorithm for SDN (CTSMA) cloud datacenter

al.¹

2022

Int. j. adv. appl. sci.

View full text Add to dashboard Cite

Multi-controller deployment in a software-defined network improves the system's stability and scalability. However, since network traffic fluctuates, it presents a new problem for balancing loads on remote controllers. Controller Adaption and Migration Decision (CAMD) and Dynamic and Adaptive Load Balancing (DALB) frameworks are developed for efficient balancing of load on the controller to solve the problem of controller overload due to dynamic network traffic. CAMD was considered to be more efficient than DALB, but when the network is more dynamic, and the incoming traffic flow is elephant flow this leads to the overall reduction in system performance. This study proposed a Convergence Time aware Switch Migration Algorithm (CTSMA) that solved the network challenge when the network is more dynamic and incoming traffic flow is more. This research developed an enhanced switch migration algorithm to address the network difficulty of dynamically changing incoming load. Because of the imbalanced distribution of load on the controllers, processing flows will have longer response times and the controllers' throughput will be reduced. Switch migration is the best method of resolving the issue. Present techniques, on the other hand, focus solely on load balancing performance while ignoring migration efficiency, thereby leading to large migration costs and excessive control overheads. To increase the load and migration efficiency of controllers, this research work developed a convergence time aware switch migration method. To find the group of underloaded controllers in the network, the improved framework looked at controller volatility and average load status. Performance comparison indicators included controller throughput, reaction time, and convergence time. According to simulation studies, CTSMA outperforms CAMD by cutting controller reaction time by roughly 6.1%, increasing controller throughput by 8.0% on average, keeping a decent load balancing rate, lowering migration costs, and maintaining the best load balancing rate.

show abstract

“…For as long as twenty years, robotized discourse acknowledgment (ASR) has been one of the essential developments in discourse union, is still habitually utilized in applications, for example, voice control and discourse to-message [21,22]. Be that as it may, the programmed acknowledgment of discourse debased by acoustic clamor and resonation is as yet viewed as an extremely testing issue [23,24].…”

Section: Introductionmentioning

confidence: 99%

Paralinguistic Speech Processing: An Overview

Abdullah

Ameen

Ahmed

et al. 2021

AJRCoS

View full text Add to dashboard Cite

Clients can adequately control PCs and create reports by speaking with the guide of innovation, discourse acknowledgment empowers records to be delivered all the more effectively in light of the fact that the program normally produces words as fast as they expressed, which is typically a lot faster than a human can compose. Discourse acknowledgment is an innovation that consequently finds the words and expressions that best match the contribution of human discourse. The most normal use of discourse acknowledgment is correspondence, where discourse acknowledgment can be utilized to create letters/messages and different reports. Point of discourse handling: - to comprehend discourse as a mechanism of correspondence, to reflect discourse for transmission and propagation; - to inspect discourse for robotized data discovery and extraction-to find some physiological highlights of the speaker. In discourse combination, there are two significant capacities. The first is the interpretation of voice to message. The second is for the content to be converted into human voice.

show abstract

SDN-based server clusters with dynamic load balancing and performance improvement

Cited by 28 publications

References 21 publications

A Scalable Virtualized Server Cluster Providing Sensor Data Storage and Web Services

A Scalable Virtualized Server Cluster Providing Sensor Data Storage and Web Services

Convergence time aware switch migration algorithm for SDN (CTSMA) cloud datacenter

Paralinguistic Speech Processing: An Overview

Contact Info

Product

Resources

About