Model-Driven Elasticity Control for Multi-Server Queues under Traffic Surges in Cloud Environments

Tadakamalla, Venkat; Menascé, Daniel A.

doi:10.1109/icac.2018.00026

Cited by 7 publications

(1 citation statement)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Due to the recent, rapid increases in the number of cloud based services and the number of internet users, there are increasingly more companies that host services through their data centers, thus causing a huge increase in the traffic volume [1]. The cloud computing environment or the internet data center consists of many servers that process internet users' requests [2]. In addition, as the amount of data is increased due to the increasing number of social network users, service providers are faced with challenges to learn how to deploy user data on many servers [3].…”

Section: Introductionmentioning

confidence: 99%

Automated Server Operation Model to Improve the Efficiency of Domain Services

Kim¹,

Park²,

Kim³

et al. 2021

ilkonline

View full text Add to dashboard Cite

While using the internet services, users tend to frequently experience the inconvenience in service due to buffering caused by overloads in the domain service server. If there are some margins for availabilities for the service server, the above issue does not occur, so it is possible to remove the inconvenience. In each time period of the domain that provides the service, the peak traffic data is collected to configure the data set, and the peak traffic data in each time period within the configured data set is summed up. After that, in this newly proposed model, the cumulative hourly peak traffic data is checked in terms of whether it exceeds the availability threshold set by an operator, and if it does, it is transferred to other pool group (pool N). The test results for the proposed model indicates that in the past, it normally took 2 hours to analyze the domain traffic data and 10 to 20 minutes to apply the analysis results, but in the newly proposed model, it takes only less than 5 minutes to analyze and apply results. System load frequency is reduced from 13 times a month to 1 time per month, and also the number of times to use the reserved server is reduced from 18 times to 2 times. The CPU margin rate is increased from 27% to 71%, which indicates that the usage rate is decreased. Overall, the operator's resource is reduced, and thus it is possible to achieve server stability as well.

show abstract

Section: Introductionmentioning

confidence: 99%