Usage Centric Green Metrics for Storage Doron Chen, Ealan Henis, Ronen Kat and Dmitry Sotnikov IBM Haifa Research Lab Most of the metrics defined today.

Slides:



Advertisements
Similar presentations
Key Metrics for Effective Storage Performance and Capacity Reporting.
Advertisements

Module – 3 Data protection – raid
SLA-Oriented Resource Provisioning for Cloud Computing
© 2009 IBM Corporation Statements of IBM future plans and directions are provided for information purposes only. Plans and direction are subject to change.
Enhanced Availability With RAID CC5493/7493. RAID Redundant Array of Independent Disks RAID is implemented to improve: –IO throughput (speed) and –Availability.
© 2009 EMC Corporation. All rights reserved. EMC Proven Professional The #1 Certification Program in the information storage and management industry Data.
2P13 Week 11. A+ Guide to Managing and Maintaining your PC, 6e2 RAID Controllers Redundant Array of Independent (or Inexpensive) Disks Level 0 -- Striped.
REDUNDANT ARRAY OF INEXPENSIVE DISCS RAID. What is RAID ? RAID is an acronym for Redundant Array of Independent Drives (or Disks), also known as Redundant.
IT Equipment Efficiency Peter Rumsey, Rumsey Engineers.
An Adaptable Benchmark for MPFS Performance Testing A Master Thesis Presentation Yubing Wang Advisor: Prof. Mark Claypool.
1 stdchk : A Checkpoint Storage System for Desktop Grid Computing Matei Ripeanu – UBC Sudharshan S. Vazhkudai – ORNL Abdullah Gharaibeh – UBC The University.
Peer-to-peer archival data trading Brian Cooper and Hector Garcia-Molina Stanford University.
© 2009 IBM Corporation Statements of IBM future plans and directions are provided for information purposes only. Plans and direction are subject to change.
Data Storage Willis Kim 14 May Types of storages Direct Attached Storage – storage hardware that connects to a single server Direct Attached Storage.
Performance Acceleration Module Tech ONTAP Live August 6 th, 2008.
1© Copyright 2013 EMC Corporation. All rights reserved. Characterization of Incremental Data Changes for Efficient Data Protection Hyong Shim, Philip Shilane,
CHAPTER OVERVIEW SECTION 5.1 – MIS INFRASTRUCTURE
“Five minute rule ten years later and other computer storage rules of thumb” Authors: Jim Gray, Goetz Graefe Reviewed by: Nagapramod Mandagere Biplob Debnath.
11 Capacity Planning Methodologies / Reporting for Storage Space and SAN Port Usage Bob Davis EMC Technical Consultant.
Disk and Tape Square Off Again Tape Remains King of Hill with LTO-4 Presented by Heba Saadeldeen.
PRESENTATION TITLE GOES HERE SNIA Emerald™ Training June 24-27, 2013 SNIA Emerald TM Training SNIA Emerald Power Efficiency Measurement Specification,
Two or more disks Capacity is the same as the total capacity of the drives in the array No fault tolerance-risk of data loss is proportional to the number.
I/O – Chapter 8 Introduction Disk Storage and Dependability – 8.2 Buses and other connectors – 8.4 I/O performance measures – 8.6.
PARAID: The Gear-Shifting Power-Aware RAID Charles Weddle, Mathew Oldham, An-I Andy Wang – Florida State University Peter Reiher – University of California,
Exploiting Flash for Energy Efficient Disk Arrays Shimin Chen (Intel Labs) Panos K. Chrysanthis (University of Pittsburgh) Alexandros Labrinidis (University.
Nexenta Proprietary Global Leader in Software Defined Storage Nexenta Technical Sales Professional (NTSP) COURSE CONTENT.
PrimeEnergyIT. The Project PrimeEnergyIT charter Market development for energy efficient central IT equipment Hardware and service based energy efficiency.
Storage Management in Virtualized Cloud Environments Sankaran Sivathanu, Ling Liu, Mei Yiduo and Xing Pu Student Workshop on Frontiers of Cloud Computing,
Improving Disk Latency and Throughput with VMware Presented by Raxco Software, Inc. March 11, 2011.
Hadoop Hardware Infrastructure considerations ©2013 OpalSoft Big Data.
C OMPUTER O RGANIZATION AND D ESIGN The Hardware/Software Interface 5 th Edition Chapter 1 Computer Abstractions and Technology Sections 1.5 – 1.11.
Tag line, tag line Power Management in Storage Systems Kaladhar Voruganti Technical Director CTO Office, Sunnyvale June 12, 2009.
Oracle Advanced Compression – Reduce Storage, Reduce Costs, Increase Performance Session: S Gregg Christman -- Senior Product Manager Vineet Marwah.
Your university or experiment logo here Caitriana Nicholson University of Glasgow Dynamic Data Replication in LCG 2008.
Eneryg Efficiency for MapReduce Workloads: An Indepth Study Boliang Feng Renmin University of China Dec 19.
Overview of Data Center Energy Use Bill Tschudi, LBNL
MSN 数学媒体与信息存储 1/27 Zhuo Liu, Fei Wu, Xiao Qin, Changsheng Xie, Jian Zhou, and Jianzong Wang TRACER: A Trace Replay Tool to Evaluate Energy-Efficiency of.
"1"1 Introduction to Managing Data " Describe problems associated with managing large numbers of disks " List requirements for easily managing large amounts.
Data Replication and Power Consumption in Data Grids Susan V. Vrbsky, Ming Lei, Karl Smith and Jeff Byrd Department of Computer Science The University.
Price Performance Metrics CS3353. CPU Price Performance Ratio Given – Average of 6 clock cycles per instruction – Clock rating for the cpu – Number of.
NAS 2011 Liang Kai,Xiaofang Zhang, Xiao Zhang Northwestern Polytechnical University Research on Energy Consumption of General Network Storage.
Group 25 Sumin Mohanan, Zoheb.H Borbora 3/8/2011.
© 2011 IBM Corporation Sizing Guidelines Jana Jamsek ATS Europe.
Preliminary Assessment of Emerald Data SNIA Emerald Metric Analysis Working Group ENERGY STAR® Discussion November 18, 2015.
PRESENTATION TITLE GOES HERE SNIA Emerald™ COM benefits – some initial data.
PRESENTATION TITLE GOES HERE SNAI Emerald 3.0 Storage Taxonomy Patrick Stanko SNIA Emerald TM SNIA Emerald Power Efficiency Measurement Specification,
Best Available Technologies: External Storage Overview of Opportunities and Impacts November 18, 2015.
PRESENTATION TITLE GOES HERE Emerald NAS Extensions Chuck Paridon Performance Architect H-P Enterprise Data Contributed by Nick Principe – EMC, Demartek.
1.3 ON ENHANCING GridFTP AND GPFS PERFORMANCES A. Cavalli, C. Ciocca, L. dell’Agnello, T. Ferrari, D. Gregori, B. Martelli, A. Prosperini, P. Ricci, E.
Considerations for Future Storage System Metrics and Efficiency Evaluation Green TWG - Design Metrics Subgroup 11/18/2015.
Maximizing Performance – Why is the disk subsystem crucial to console performance and what’s the best disk configuration. Extending Performance – How.
Enhanced Availability With RAID CC5493/7493. RAID Redundant Array of Independent Disks RAID is implemented to improve: –IO throughput (speed) and –Availability.
1 Paolo Bianco Storage Architect Sun Microsystems An overview on Hybrid Storage Technologies.
1 © Copyright IBM Corporation /2013 The Sizer Advisor - v1.0 The Sizer Advisor Ten Elementary Advices Every Sizer Should Follow v1.0 Jorge L. Navarro.
Presenter: Yue Zhu, Linghan Zhang A Novel Approach to Improving the Efficiency of Storing and Accessing Small Files on Hadoop: a Case Study by PowerPoint.
System Storage TM © 2007 IBM Corporation IBM System Storage™ DS3000 Series Jüri Joonsaar Tartu.
Input and Output Optimization in Linux for Appropriate Resource Allocation and Management James Avery King.
Measuring Performance II and Logic Design
Getting the Most out of Scientific Computing Resources
Answer to Summary Questions
Getting the Most out of Scientific Computing Resources
Seth Pugsley, Jeffrey Jestes,
Temperature Aware Storage
IT Equipment Efficiency
-A File System for Lots of Tiny Files
IT Equipment Efficiency
Qingbo Zhu, Asim Shankar and Yuanyuan Zhou
Troubleshooting Techniques(*)
The Greening of IT November 1, 2007.
Storage Modeling for Power Estimation Ronen Kat
Presentation transcript:

Usage Centric Green Metrics for Storage Doron Chen, Ealan Henis, Ronen Kat and Dmitry Sotnikov IBM Haifa Research Lab Most of the metrics defined today for storage measure the energy efficiency potential of a resource, system or application usage, rather than the energy efficiency of the actual usage. While the efficiency potential is important, the way that the resources and systems are actually used is at least as important. Hence, we suggest to both aim at energy efficient components (as done today), as well as optimize the actual usage of the components and systems in the data center. 2. Motivation 4. Key Players for Energy Efficient Storage The U.S Environmental Protection Agency (EPA) The Green Grid The Storage Networking Industry Association (SNIA) The Storage Performance Council (SPC) The Standard Performance Evaluation Corporation (SPEC) 1. Abstract Contact: Doron Chen, Ealan Henis, Ronen Kat and Dmitry Sotnikov {cdoron,ealan With government focus and regulation on data center storage, energy efficiency and energy metrics receive more attention. Current metrics that measure potential energy efficiency of a storage resource are not enough – actual storage usage efficiency metrics are required for system characterization. Used Capacity: 90 GB Power Consumption: 4*15 W = 60 W Raw Capacity: 4*25 GB = 100 GB Used Capacity: 90 GB Power Consumption: 2*15 W = 30 W Raw Capacity: 2*25 GB = 50 GB Potential Capacity metric: 100/60=1.67 GB/W Usage Capacity metric: 135/60=2.25 GB/W Potential Capacity metric: 100/60=1.67 GB/W Usage Capacity metric: 90/60=1.5 GB/W Potential Capacity metric: 50/30=1.67 GB/W Usage Capacity metric: 90/30=3 GB/W Used Capacity: 135 GB Power Consumption: 4*15 W = 60 W Raw Capacity: 4*25 GB = 100 GB Capacity Metric: how data de-duplication shows up in the metrics 3. Goal Adopting usage centric metrics has merit for energy efficiency, and should be used at all system levels as a primary means of assessing a system’s energy efficiency. The Capacity Metric (GB/Watt) represents the energy efficiency of storing data 5. Usage Metrics versus Potential Metrics ratio between used storage space ¥ by applications (GB) and power § (Watts) ratio between raw space (GB) and power (Watts) measured during idle period ratio between predefined benchmark I/O rate (IOPS) and power (Watts) ratio between applications I/O rate (IOPS) and power (Watts) The I/O Throughput Metric (IOPS/Watt) represents the energy cost of the storage system while running the workload. The Data Transfer Throughput Metric (MBPS/Watt) similar to I/O throughput, referring to data transfer (MBPS) The two throughput metrics emphasize different aspects of the same efficiency. I/O throughput is commonly used for evaluating random workloads, and data transfer throughput is commonly used for evaluating sequential workloads ¥ Used storage space is defined as the space used by files written and stored on the storage system. § Power is measured as the average power of the storage system under typical usage, measured over a representative (long enough) time period ratio between predefined benchmark data transfer rate (MBPS) and power (Watts) ratio between applications data transfer rate (MBPS) and power (Watts) 90% of space used 45% of space used For example, consider two 300 GB disks that use 10.6 Watts each, the efficiency potential value is equal for both. However, the space usage of the disks may be different, and the lower usage disk may be replaced with a smaller, lower power consumption disk – saving more than 10% energy. Fault Tolerance: 2 disks Fault Tolerance: 1 disk RAID 6 RAID 5 It is obvious that the RAID 5 configuration is more space efficient than RAID 6, and if we need to store the same user capacity in the RAID 6 system, we will use more physical disks. If, for example, each disk has 100 GB, to store 300 GB of user data at RAID 5 we need only 4 disks; with RAID 6 we need 5. The potential efficiency metrics will evaluate both configuration to have the same level. The usage efficiency metrics will evaluate the RAID 5 configuration as better. Note that our usage efficiency metrics evaluate only usage; the system administrator needs to consider other requirements such as fault tolerance levels and others. Base system Using de-duplication: can store more data with same capacity Using de-duplication: can store same data with less capacity