Holistic Virtual Machine Scheduling in Cloud Datacenters towards Minimizing Total Energy

Abstract

Abstract-Energy consumed by Cloud datacenters has dramatically increased, driven by rapid uptake of applications and services globally provisioned through virtualization. By applying energy-aware virtual machine scheduling, Cloud providers are able to achieve enhanced energy efficiency and reduced operation cost. Energy consumption of datacenters consists of computing energy and cooling energy. However, due to the complexity of energy and thermal modeling of realistic Cloud datacenter operation, traditional approaches are unable to provide a comprehensive in-depth solution for virtual machine scheduling which encompasses both computing and cooling energy. This paper addresses this challenge by presenting an elaborate thermal model that analyzes the temperature distribution of airflow and server CPU. We propose GRANITE -a holistic virtual machine scheduling algorithm capable of minimizing total datacenter energy consumption. The algorithm is evaluated against other existing workload scheduling algorithms MaxUtil, TASA, IQR and Random using real Cloud workload characteristics extracted from Google datacenter tracelog. Results demonstrate that GRANITE consumes 4.3% -43.6% less total energy in comparison to the state-of-the-art, and reduces the probability of critical temperature violation by 99.2% with 0.17% SLA violation rate as the performance penalty.

ETAS: Energy and thermal‐aware dynamic virtual machine consolidation in cloud data center with proactive hotspot mitigation

“…In a similar way, Li et al 25 have proposed scheduling algorithm for holistic energy minimization of computing and cooling system in cloud data centers. The authors have proposed a greedy heuristic scheduling algorithm GRANITE that balances workload after fixed scheduling interval.…”

“…*https://aws.amazon.com/ec2/ • GRANITE: Greedy VM scheduling algorithm to minimize holistic energy in cloud data center proposed in the work of Li et al 25 This policy dynamically migrates VM to balance workload based on a certain temperature threshold.…”

Heuristics and metaheuristics for dynamic management of computing and cooling energy in cloud data centers

“…Their work presents the data center as a distributed cyberphysical system in which the energy of both IT and cooling is the minimization objective. Li et al used a greedy scheduling algorithm, considering the temperature distribution airflow to minimize the energy consumption of IT and cooling jointly. Current research in the area of joint energy and thermal aware strategies consider fan power and leakage due to temperature or cooling units' power but do not take into account the combined effects between them.…”

A Framework and Task Allocation Analysis for Infrastructure Independent Energy-Efficient Scheduling in Cloud Data Centers

“…However, this approach leads to the formation of hot spots within data centers resulting in increased cooling energy [3], [4]. This trade-off is inherently difficult to understand and leads to suboptimal reduction in data center energy [5]. Second, analytical modeling of energy consumption within the data center is a substantive task, especially when modeling cooling energy which captures various fluid mechanical and thermodynamic aspects of the system.…”

“…Second, analytical modeling of energy consumption within the data center is a substantive task, especially when modeling cooling energy which captures various fluid mechanical and thermodynamic aspects of the system. Lastly, existing approaches that directly model total cooling energy are typically restricted to a certain class of data center architecture [5], [6], [7]. In particular, cooling models heavily depend on the data center geometry (i.e.…”