Sensitivity Based Power Management of Enterprise Storage Systems

Sankar, Sriram; Gurumurthi, Sudhanva; Stan, Mircea R.

doi:10.1109/mascot.2008.4770578

Cited by 8 publications

(5 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Multiple references in circuit design [21,22], disk drive design [23,24] and smartphone design [25] use energy-delay optimization techniques to establish optimal performance-per-watt. For a given energy constraint and SLA, we optimize 'energy*delay' to find an optimal operating point.…”

Section: Energy-delay Characterization Methodologymentioning

confidence: 99%

“…Zhang et al [23] demonstrated the applicability of sensitivity-based optimization for determining the optimal settings of the design-time parameters of disk drives. A dynamic energy-delay optimization technique was developed for energy reduction in disk drives [24]. Smartphone design [25] also uses energy-delay optimization to extend battery life.…”

Section: Energy Delay Optimizationmentioning

confidence: 99%

See 1 more Smart Citation

Energy-delay based provisioning for large datacenters: an energy-efficient and cost optimal approach

Sankar

Vaid

Rogers

2011

Proceedings of the 2nd ACM/SPEC International Conference on Performance Engineering

Self Cite

View full text Add to dashboard Cite

It is challenging to determine the optimum number of servers required to provision for large online applications because of the conflicting mandates of (a) achieving peak performance needs and (b) minimizing unused datacenter power and capacity. Since Online Services application loads are unpredictable, datacenter operators often conservatively provision for maximum power utilization by characterizing workloads for peak load performance. In contrast, we aim to optimize the service capacity per total cost of ownership (TCO) of an Online Service datacenter deployment by characterizing the energy-delay properties of large scale datacenter workloads. We show that the peak load performance is not the energy efficient point of operation for most applications. We choose two industry-strength workloads (Internet Search and D-Process) and analyze their energy-delay behavior under varying loads. We then calculate the optimal operating point for the specific large-scale application and provision datacenter energy and capacity based on the energydelay curves.In contrast to workload-based peak power provisioning, we show a 7% benefit in Service Capacity-per-TCO-dollar for the energy-delay characterization methodology in our cost analysis for Online Services Applications.

show abstract

Section: Energy-delay Characterization Methodologymentioning

confidence: 99%

Section: Energy Delay Optimizationmentioning

confidence: 99%

Energy-delay based provisioning for large datacenters: an energy-efficient and cost optimal approach

Sankar

Vaid

Rogers

2011

Proceedings of the 2nd ACM/SPEC International Conference on Performance Engineering

Self Cite

View full text Add to dashboard Cite

show abstract

“…Many approaches have recently been proposed for efficient power saving in storage systems [Chase et al 2001;Ganesh et al 2007;Lim et al 2010;Sankar et al 2008;Thereska et al 2009;Verma et al 2010]. The proposed techniques include adaptive resource provisioning [Chase et al 2001]; disk access prediction with log-structure file system [Ganesh et al 2007]; disk spin-down with write off-loading ]; power-aware layout of data across servers and racks with predictive gear scheduling [Thereska et al 2009]; storage virtualization layer optimization [Verma et al 2010]; and dynamic provisioning with a simple integral control [Lim et al 2010].…”

Section: Power Managementmentioning

confidence: 99%

“…However, this approach risks a significant wastage of energy during periods of low I/O load. To address this challenge, prior approaches on power management focus on modulating the number of servers [Ganesh et al 2007;Lim et al 2010;Sankar et al 2008;Thereska et al 2009;Verma et al 2010] without addressing how to migrate their data in a costeffective manner. In contrast, we observe that our proposed dynamic reconfiguration technique can be used for power management, and Ursa performs power consolidation in a topology-aware manner over the granularity of a few hours, alleviating concerns of limited disk spin-down cycles.…”

Section: Introductionmentioning

confidence: 99%

Ursa

2013

View full text Add to dashboard Cite

Enterprise and cloud data centers are comprised of tens of thousands of servers providing petabytes of storage to a large number of users and applications. At such a scale, these storage systems face two key challenges: (1) hot-spots due to the dynamic popularity of stored objects; and (2) high operational costs due to power and cooling. Existing storage solutions, however, are unsuitable to address these challenges because of the large number of servers and data objects. This article describes the design, implementation, and evaluation of Ursa, a system that scales to a large number of storage nodes and objects, and aims to minimize latency and bandwidth costs during system reconfiguration. Toward this goal, Ursa formulates an optimization problem that selects a subset of objects from hot-spot servers and performs topology-aware migration to minimize reconfiguration costs. As exact optimization is computationally expensive, we devise scalable approximation techniques for node selection and efficient divide-and-conquer computation. We also show that the same dynamic reconfiguration techniques can be leveraged to reduce power costs by dynamically migrating data off under-utilized nodes, and powering up servers neighboring existing hot-spots to reduce reconfiguration costs. Our evaluation shows that Ursa achieves cost-effective load management, is time-responsive in computing placement decisions (e.g., about two minutes for 10K nodes and 10M objects), and provides power savings of 15%-37%.

show abstract

“…This work considered static measurement of device performance in deciding on migration as opposed to our method of employing dynamically measured response times. Dynamic measures are employed in managing power consumption in storage systems in [12]. Flash memory is being used as a cache to improve disk drive performance [e.g., 25].…”

Section: Related Workmentioning

confidence: 99%

Managing storage space in a flash and disk hybrid storage system

Reddy

2009

2009 IEEE International Symposium on Modeling, Analysis &Amp; Simulation of Computer and Telecommunication Systems

View full text Add to dashboard Cite

This paper considers the problem of efficiently managing storage space in a hybrid storage system employing flash and disk drives. The flash and disk drives exhibit different performance characteristics of read and write behavior. We propose a technique for balancing the workload properties across flash and disk drives in such a hybrid storage system. We consider various alternatives for managing the storage space in such a hybrid system and show that the proposed technique improves performance in diverse scenarios. The presented approach automatically and transparently manages migration of data blocks among flash and disk drives based on their access patterns. We implemented a prototype storage system employing our ideas. This paper presents the design and an evaluation of the proposed approach through realistic experiments.

show abstract

Sensitivity Based Power Management of Enterprise Storage Systems

Abstract: Energy-efficiency is a key requirement in data centers

Cited by 8 publications

References 16 publications

Energy-delay based provisioning for large datacenters: an energy-efficient and cost optimal approach

Energy-delay based provisioning for large datacenters: an energy-efficient and cost optimal approach

Ursa

Managing storage space in a flash and disk hybrid storage system

Contact Info

Product

Resources

About