Cloud Service Differentiation in Overbooked Data Centers

Tomás, Luis; Tordsson, Johan

doi:10.1109/ucc.2014.83

Cited by 8 publications

(3 citation statements)

References 26 publications

(31 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Os autores foram capazes de preservar o consumo de energia elétrica do provedor da NC. Tordsson [Tomás and Tordsson 2014] propõe uma nova configurac ¸ão de NC para habilitar o mapeamento de servic ¸os críticos de um cliente, referente ao aperfeic ¸oamento da performance de uso de núcleos de CPU físico. Idealmente, a identificac ¸ão de overbooking de recursos deve ser feita pelo lado do cliente devido ao conflito de interesses.…”

Section: Trabalhos Relacionadosunclassified

Detecção de Overbooking em Aplicações Baseadas em Docker Através de Aprendizagem de Máquina

Horchulhack

Viegas

Santin

et al. 2022

Anais Estendidos Do XL Simpósio Brasileiro De Redes De Computadores E Sistemas Distribuídos (SBRC Estendido 2022)

View full text Add to dashboard Cite

O artigo propõe um modelo de aprendizado de máquina para detectar ambientes Kubernetes com overbook de recursos em um contêiner do Docker. As métricas do aplicativo e do sistema foram coletadas continuamente, as quais foram usadas como entrada para o modelo para identificar interferência causada por multi-tenancy. Os experimentos foram executados em um cluster Kubernetes, com um aplicativo de Big Data baseado em contêiner, o que mostrou que o modelo pode detectar overbooking de recursos com precisões de até 98% em um cenário que pode causar degradação no desempenho do aplicativo, com taxas de overbooking de até 1,2.

show abstract

Section: Trabalhos Relacionadosunclassified

Detecção de Overbooking em Aplicações Baseadas em Docker Através de Aprendizagem de Máquina

Horchulhack

Viegas

Santin

et al. 2022

Anais Estendidos Do XL Simpósio Brasileiro De Redes De Computadores E Sistemas Distribuídos (SBRC Estendido 2022)

View full text Add to dashboard Cite

show abstract

“…Our previous framework either relayed either on the KVM scheduler to perform the physical CPU (pcpu) sharing among VMs [11], or made use of the KVM pinning functionality to provide some isolation among VMs, with quality of service (QoS) differentiation purposes [17]. In the latter, we present a method to isolate some VMs from others (and consequently limiting their impact on each other) inside an overbooked server by efficiently using KVM virtual to physical CPU pinning.…”

Section: Background and Related Workmentioning

confidence: 99%

Reducing Noisy-Neighbor Impact with a Fuzzy Affinity-Aware Scheduler

Tomás

Vázquez²,

Tordsson³

et al. 2015

2015 International Conference on Cloud and Autonomic Computing

Self Cite

View full text Add to dashboard Cite

Abstract-Overbooking techniques have been proven efficient to increase overall utilization of cloud datacenters. However, overbooking may also degrade applications performance as (at least) some applications need to share physical resources such as CPU or memory. Consequently, interference may increase among the virtual machines that share resources, the so called noisy neighbors effect. We present an affinity-aware scheduler to reduce the impact of such interference. A fuzzy logic engine accounts for the uncertainty in these environments and estimates which CPU cores are currently more suitable for each incoming application. This helps the scheduler make virtual machine to physical resource mapping decisions, also known as vcpu pinning. An experimental evaluation based on a combination of interactive services and batch applications confirms that our affinity-aware fuzzy scheduler reduces the interference among applications, enabling more predictable performance and consequently safer overbooking.

show abstract

“…Over provisioning increases the management cost for a cloud application, but guarantees user satisfaction. The second strategy is provisioning according to expectations and handling capacity shortages via user experience degradation [7,9,27,34,41], or via approximate computing [22,37,40]. Generally speaking, these ways of handling capacity shortages are typically clustered under the umbrella of graceful degradation.…”

Section: Introductionmentioning

confidence: 99%

Cloud Application Predictability through Integrated Load-Balancing and Service Time Control

Nylander

Andrén

Årzén

et al. 2018

2018 IEEE International Conference on Autonomic Computing (ICAC)

View full text Add to dashboard Cite

Cloud computing provides the illusion of infinite capacity to application developers. However, data center provisioning is complex and it is still necessary to handle the risk of capacity shortages. To handle capacity shortages, graceful degradation techniques sacrifice user experience for predictability. In all these cases, the decision making policy that determines the degradation interferes with other decisions happening at the infrastructure level, like load-balancing choices. Here, we reconcile the two approaches, developing a load-balancing strategy that also handles capacity shortages and graceful degradation when necessary. The proposal is based on a sound control-theoretical approach. The design of the approach avoids the pitfalls of interfering control decisions. We describe the technique and provide evidence that it allows us to achieve higher performance in terms of emergency management and user experience.

show abstract

Cloud Service Differentiation in Overbooked Data Centers

Cited by 8 publications

References 26 publications

Detecção de Overbooking em Aplicações Baseadas em Docker Através de Aprendizagem de Máquina

Detecção de Overbooking em Aplicações Baseadas em Docker Através de Aprendizagem de Máquina

Reducing Noisy-Neighbor Impact with a Fuzzy Affinity-Aware Scheduler

Cloud Application Predictability through Integrated Load-Balancing and Service Time Control

Contact Info

Product

Resources

About