PAP: power aware prediction based framework to reduce disk energy consumption
- PDF / 2,357,086 Bytes
- 18 Pages / 595.276 x 790.866 pts Page_size
- 69 Downloads / 138 Views
(0123456789().,-volV)(0123456789(). ,- volV)
PAP: power aware prediction based framework to reduce disk energy consumption Sumedha Arora1 • Anju Bala1 Received: 2 October 2019 / Revised: 16 January 2020 / Accepted: 23 February 2020 Ó Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract Disk-based storage subsystems account for a significant portion of the energy consumption in both low and high end servers. Therefore, there is a dire need to reduce the server power consumption of the hard disks. In this work, the poweraware framework has been proposed, which efficiently switches the disk into standby, active and idle states, leading to the least power consumption. Firstly, the trace of a real-world application has been generated and processed. The frequently used queries from the trace have been analyzed and prefetched in SSD cache using the data placement policy which lead to 78.5% cache hits. Subsequently, the idle time threshold policy has been executed, which regularly monitors and compares the disk idle time with its threshold value. Later, the request arrival threshold policy predicts the breakeven time using the ensemble machine learning model, which yields 87% accuracy with 3.5% average error rate. Only upon exceeding the threshold values, the disk would smartly be placed in the standby mode; otherwise, it would remain in the idle state to avoid the high power spins in case of frequent requests. Finally, the experimental results have been validated with the existing benchmarks using SSD as a cache, which leads to 75% average power savings. Keywords Energy consumption Prediction Power Arrival Request rate Disk
1 Introduction Energy has become a critical resource in modern computing systems, which presents challenges to the traditional storage systems [1]. The energy consumed by the storage sub-system surpasses all other sub-components present in the server. The disks in high-end servers are responsible for the high power consumption [2]. According to the report recently published by the Natural Resources Defense Council (NRDC), the energy consumed by the data centers is largely due to the spinning of the hard disk for storing and processing the data. A hard disk platter requires high power to spin 24/7/365 at 7000 or 10,000 RPMs [3]. Hence, the server’s power consumption can be decreased by obtaining an in-depth understanding of the & Sumedha Arora [email protected] Anju Bala [email protected] 1
Computer Science and Engineering Department, Patiala, India
spinning trends for hard drives [4]. A disk can run in highpower or low-power modes according to the data request or can even be shut off so that it does not consume power at a time of no request. In a large cluster of disks, only a fraction of it is accessed at any point of time. The rest of the disk could potentially be switched to a low-power mode [5]. Numerous researchers have replaced Hard Disk Drives (HDDs) with Solid-State Drives (SSDs) [6] to achieve energy efficiency. The SSDs can be used as a cache to p
Data Loading...
