Dynamic Power- and Failure-Aware Cloud Resources Allocation for Sets of Independent Tasks

Sampaio, Altino M.; Barbosa, Jorge G.

doi:10.1109/ic2e.2013.16

Cited by 19 publications

(19 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As a result of this lack of empirical study, a sizable body of current Cloud dependability mechanisms and workload characterization research is derived from analysis of other distributed systems [8][9][10] or incorporates theoretical values [11][12][13][14]. While such work is relevant to enhancing Cloud dependability mechanisms and workload characterization, it comes with significant limitations.…”

Section: Introductionmentioning

confidence: 99%

An Empirical Failure-Analysis of a Large-Scale Cloud Computing Environment

Garraghan

Townend

2014

2014 IEEE 15th International Symposium on High-Assurance Systems Engineering

View full text Add to dashboard Cite

Abstract-Cloud computing research is in great need of statistical parameters derived from the analysis of real-world systems. One aspect of this is the failure characteristics of Cloud environments composed of workloads and servers; currently, few metrics are available that quantify failure and repair times of workloads and servers at a large-scale. Workload metrics in particular are critical for characterizing and modeling accurate workload behavior, enabling more realistic workload simulation and failure scenarios of systems. This paper presents the analysis of failure data of a large-scale production Cloud environment (consisting of over 12,500 servers), and includes a study of failure and repair times and characteristics for both Cloud workloads and servers. Our results show that failure characteristics for workload and servers are highly variable and that production Cloud workloads can be accurately modeled by a Gamma distribution. Repair times range between 30 seconds to 4 days, and 25 minutes to 8 days, for workloads and servers respectively.

show abstract

Section: Introductionmentioning

confidence: 99%

An Empirical Failure-Analysis of a Large-Scale Cloud Computing Environment

Garraghan

Townend

2014

2014 IEEE 15th International Symposium on High-Assurance Systems Engineering

View full text Add to dashboard Cite

show abstract

“…This work extends our previous work [2] by considering the slowdown β i (t) experienced by task t in node i due to contention in shared on-chip resources, as expressed in (4).…”

Section: B Performance Interference Mitigation Objectivementioning

confidence: 60%

“…Our strategy aims at improving energy efficiency, while maximizing the rate of completed jobs by their deadlines. This work represents an extension of [2], and the main contributions are: (i) a mechanism to determine the effective slowdown of tasks, based on noisy data samples provided by state-of-the-art slowdown meters; (ii) a scheduling algorithm that constructs interference-and power-aware virtual clusters. The aim is to build virtual clusters able to successfully accomplish applications QoS requirements (execute the tasks by their deadlines), with less energy consumption.…”

Section: Introductionmentioning

confidence: 99%

Estimating Effective Slowdown of Tasks in Energy-Aware Clouds

Sampaio

Barbosa

2014

2014 IEEE International Symposium on Parallel and Distributed Processing With Applications

View full text Add to dashboard Cite

Abstract-Consolidation consists in scheduling multiple virtual machines onto fewer servers in order to improve resource utilization and to reduce operational costs due to power consumption. However, virtualization technologies do not offer performance isolation, causing applications' slowdown. In this work, we propose a performance enforcing mechanism, composed of a slowdown estimator, and a interference-and power-aware scheduling algorithm. The slowdown estimator determines, based on noisy slowdown data samples obtained from state-of-the-art slowdown meters, if tasks will complete within their deadlines, invoking the scheduling algorithm if needed. When invoked, the scheduling algorithm builds performance and power aware virtual clusters to successfully execute the tasks. We conduct simulations injecting synthetic jobs which characteristics follow the last version of the Google Cloud tracelogs. The results indicate that our strategy can be efficiently integrated with state-of-the-art slowdown meters to fulfil contracted SLAs in real-world environments, while reducing operational costs in about 12%.

show abstract

“…Sampaio et al [5] targets to improve cloud infrastructure power efficiency using dynamic algorithms which map physical machines to virtual machines. They use a proactive fault tolerance technique to identify failures in systems.…”

Section: Related Researchesmentioning

confidence: 99%

“…Cloud is a group of physical machines pretending to be one computing environment [4].User see cloud as an illusion of unlimited computing resources [5]. But the main challenge is to manage the variability and heterogeneity of application requirement [6].…”

Section: Introductionmentioning

confidence: 99%

Identifying Overloaded Servers and Managing Dynamic Placement of Virtual machines in Cloud

Rawshan¹,

Khan²,

Imran³

2016

IJCA

View full text Add to dashboard Cite

Cloud computing is becoming one of the most popular commercial infrastructure due to its little maintenance expense and on demand resource utilization. Cloud computing possesses many kinds of technical challenges such as fault tolerance, reliability, availability, integrity etc. due to its complex and distributed nature. But the main problem related to all those is overload incurred by Virtual Machines (VM). So, load balancing is one of the most significant issues that can help to gain rapid performance of cloud infrastructure. This research proposes algorithms for detecting failed servers due to overloaded VMs. The failure detection algorithm checks server status after a predefined time interval. This algorithm gives proactive technique to deal with overloaded VMs. When any failure in the server is found, the resource balancing algorithm migrates its VMs to an adequate healthy Physical Machine (PM). To distribute workload evenly, the resource utilization skew is measured. This VM to PM mapping is done in a way that every PM will do almost equal amount of work.

show abstract

Dynamic Power- and Failure-Aware Cloud Resources Allocation for Sets of Independent Tasks

Cited by 19 publications

References 24 publications

An Empirical Failure-Analysis of a Large-Scale Cloud Computing Environment

An Empirical Failure-Analysis of a Large-Scale Cloud Computing Environment

Estimating Effective Slowdown of Tasks in Energy-Aware Clouds

Identifying Overloaded Servers and Managing Dynamic Placement of Virtual machines in Cloud

Contact Info

Product

Resources

About