1、Server EfficiencyMetrics for Computer Serversand StorageServer Efficiency: Metrics for Computer Servers and Storage is authored by ASHRAE Technical Committee (TC) 9.9, Mission Critical Facilities, Data Centers, Technology Spaces and Electronic Equipment. ASHRAE TC 9.9 is composed of a wide range of
2、industry representatives, including but not limited to equipment manufacturers, consulting engineers, data center operators, academia, testing laboratories, and government officials who are all committed to increasing and sharing the body of knowledge related to data centers.Any updates/errata to th
3、is publication will be posted on the ASHRAE website at www.ashrae.org/publicationupdates.For more information on the ASHRAE Datacom Series, visitwww.ashrae.org/datacenterefficiency.For more information on ASHRAE TC 9.9, visit http:/tc99.ashraetcs.org/.Server EfficiencyMetrics for Computer Serversand
4、 StorageASHRAE Datacom SeriesBook 12Klaus-Dieter Lange and Karl Huppler, EditorsAtlantaISBN 978-1-936504-91-6 2015 ASHRAE1791 Tullie Circle, NEAtlanta, GA 30329www.ashrae.orgAll rights reserved.Printed in the United States of AmericaCover image by Joe Lombardo of DLB Associates.ASHRAE is a registere
5、d trademark in the U.S. Patent and Trademark Office,owned by the American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc.ASHRAE has compiled this publication with care, but ASHRAE has not investigated, and ASHRAE expressly disclaims any duty to investigate, any product, servic
6、e, process, procedure, design, or the like that may be described herein. The appearance of any technical data or editorial material in this publi-cation does not constitute endorsement, warranty, or guaranty by ASHRAE of any product, service, process, procedure, design, or the like. ASHRAE does not
7、warrant that the information in the publica-tion is free of errors, and ASHRAE does not necessarily agree with any statement or opinion in this publication. The entire risk of the use of any information in this publication is assumed by the user.No part of this publication may be reproduced without
8、permission in writing from ASHRAE, except by a reviewer who may quote brief passages or reproduce illustrations in a review with appropriate credit, nor may any part of this publication be reproduced, stored in a retrieval system, or transmitted in any way or by any meanselectronic, photocopying, re
9、cording, or otherwithout permission in writing from ASHRAE. Requests for permission should be submitted at www.ashrae.org/permissions.Library of Congress Cataloging-in-Publication DataServer efficiency-metrics for computer servers and storage.pages cm. - (Ashrae datacom series ; book 12)Includes bib
10、liographical references.Summary: “Presents tools and benchmarks for gauging the energy efficiency of a computer server or storage subsystem, and provides in-depth case studies on the application of these metrics“- Provided by publisher.ISBN 978-1-936504-91-6 (softcover)1. Data processing service cen
11、ters-Energy conservation. 2. Storage area networks (Computer networks)-Energy conservation. 3. Web servers-Energy conservation. 4. Web servers-Energy consumption-Measurement. I. ASHRAE (Firm)TJ163.5.D38S47 2015621.392-dc232015006616ASHRAE STAFF SPECIAL PUBLICATIONS Mark S. Owen, Editor/Group Manager
12、 of Handbook and Special PublicationsCindy Sheffield Michaels, Managing EditorJames Madison Walker, Managing EditorSarah Boyle, Assistant EditorLauren Ramsdell, Editorial AssistantMichshell Phillips, Editorial CoordinatorPUBLISHING SERVICES David Soltis, Group Manager of Publishing Services and Elec
13、tronic CommunicationsJayne Jackson, Publication Traffic AdministratorTracy Becker, Graphics SpecialistPUBLISHER W. Stephen ComstockKlaus-Dieter LangeChapters 1, 3, 4, 5 and 9Hewlett-Packard CompanyDavid SchmidtChapter 4Hewlett-Packard CompanyDavid MorseChapter 6VMwareMatthais Koechl Chapter 7IBMHeik
14、o GerwensChapter 7SAPKarl HupplerChapters 1, 2, and 8IBM (Retired)Paul CaoChapter 5Hewlett-Packard CompanyRebecca GriderChapter 6VMwareTobias KutningChapter 7SAPJ. Carlos PrattChapter 8IBMContributorsEditorsKlaus-Dieter Lange Karl HupplerHewlett-Packard Company IBM (Retired)Chapter AuthorsJay Dietri
15、chChapter 10IBMThe individual authors and editors of this book are deserving of recognition. Brief overviews of their qualifications are included in order to highlight the high levels of expertise being shared with the reader.Klaus-Dieter Lange is a master technologist at the Hewlett-Packard Company
16、 (HP), where he has worked since 1998. He began his professional career after receiv-ing his bachelors degrees from the University of Applied Sciences, Lbeck, Germany, and the Milwaukee School of Engineering in 1997. Klaus spent his career working on performance evaluation, workload characterization
17、, benchmark devel-opment, and computer systems energy efficiency. He has been active in several of the Standard Performance Evaluation Corporation (SPEC) committees since 2005 and serves as a director of its board. Klaus is the founding chair of the SPECpower Committee, which developed the SPEC Powe
18、r and Performance Methodology, the SPECpower_ssj2008 benchmark, the SPEC PTDaemon, and the Server Efficiency Rating Tool (SERT). Klaus received several of SPECs SPECtacular Awards over the years, as well as the SPEC Presidential Award in 2013. He is also actively involved in the SPEC research commun
19、ity and the International Conference of Performance Engineering.Karl Huppler recently retired from IBM, where he dedicated much of his career to measures of performance and computer energy efficiency. Karl received his bachelors degree from the University of WisconsinMadison. He served as IBMs prima
20、ry representative to the Transaction Processing Performance Council (TPC) for many years and was elected to serve as the chairman of the TPC for five years. Karl has also served as both secretary and vice-chair of the SPEC Power subcom-mittee and continues to serve that committee as a supporting con
21、tributor. He is also on the advisory board for the Linked Data Benchmark Council, a relatively new benchmark consortium. Karl has received several SPECtacular Awards through SPEC, has been a five-time invited speaker at the TPCs annual technology confer-ences, and has authored or coauthored a number
22、 of papers focused on performance and efficiency measures for computers.Paul Cao is a master technologist at HP. He has a doctoral degree in physics from Rice University, Houston. During his 16-year career with HP, Paul has contrib-uted in the fields of database and computer systems performance, opt
23、imization, and modeling. These contributions have a direct and positive impact to the technical and performance benchmark leadership of multiple generations of HP servers and stor-age products. Paul currently chairs the TPC-Energy subcommittee within the TPC.About the Authorsviii About the AuthorsJa
24、y Dietrich is an IBM distinguished engineer on the corporate environmental affairs staff, with global responsibilities for climate stewardship and energy manage-ment. Since 1982, he has been with IBM, where he develops corporate policies, strat-egies, and goals and objectives in these areas and serv
25、es as an IBM representative to external government, nongovernmental, and industry groups on the subjects of product energy-efficiency attributes; management and reduction of operational energy use and greenhouse gas (GHG) emissions in data center, manufacturing, and office environments; energy and c
26、limate change programs; and work on GHG management activities in the supply chain. He works closely with the product devel-opment and data center teams to implement energy-efficiency initiatives and proac-tively manage product and data center efficiency requirements. Jay has a bachelors degree in ci
27、vil engineering and a masters degree in environmental engineering.Heiko Gerwens is the leading performance expert for SAP financial service applications. He joined SAP in 1997 and has worked on the performance and data management team. During this time he has been involved in the development of many
28、 core insurance applications as well as in customer projects implementing these applications. He supports customers, partners, consulting, and the field organization concerning all aspects of performance, scalability, and sizing.Rebecca Grider is a member of the VMware performance engineering group.
29、 She contributes to VMmark development and uses the benchmark to measure hard-ware and virtualization platform performance. Rebecca has a masters degree in computer science from the University of Texas at Austin.Matthias Koechl is a certified senior architect with IBM, Open Group, and holds an advan
30、ced certificate in marketing management practice from the Chartered Institute of Marketing, UK. In his current role at the IBM SAP International Compe-tence Center he is in charge of worldwide sales and partner enablement for new tech-nologies and products of IBM Systems and Technology Group as well
31、 as SAP SE. Given his long experience, Matthias is knowledgeable about cross-platform infra-structure architectures, their sizing, and operations. He represented IBM in the SAP benchmark workgroup during definition of the SAP power benchmark.Tobias Kutning joined SAP in 2001. Until 2007, he worked a
32、s a database support consultant in SAP Service and Support with a focus on IBM DB2 LUW. In 2008, he joined SAP IT as a senior database administrator. Since 2010, he has worked on the performance and scalability team, responsible for SAP Standard Application Benchmark Product management.David Morse i
33、s a member of the VMware Performance Engineering Group. He has 15 years of benchmarking experience at Dell Inc., NCR Corp, and VMware. He has led benchmarking teams that were responsible for numerous benchmark lead-ership positions. Since 1999, David has held a variety of roles within the SPEC orga-
34、nization, including chairing the open systems steering committee and serving on the board of directors. He has a bachelors degree in computer engineering from the University of South Carolina and is a Red Hat Certified Engineer.David Schmidt is an expert systems engineer in the HP Servers division a
35、t Hewlett-Packard. He earned bachelors and masters degrees in physics from the University of Houston. His primary area of focus for the last 11 years has been Server EfficiencyMetrics for Computer Servers and Storage ixperformance analysis of computer servers using industry-standard benchmarks. He h
36、as published hundreds of results using the STREAM, LINPACK, SPEC CPU2000, SPEC CPU2006, SPECjbb2005, SPECpower_ssj2008, SPECvirt_sc2010, and SPECvirt_sc2013 benchmarks. David has been active as a technical representative for HP to SPEC for eight years, contributing to the development of several benc
37、h-marks, including SPEC CPU2006 and SPECvirt_sc2013. He has received multiple SPEC SPECtacular Awards for his contributions. David is co-author of the “SPECvirt_sc2010Driving Virtualization Innovation” poster paper from the 2012 third ACM/SPEC International Conference on Performance Engineering. He
38、is currently the chair of the SPEC Open Systems Groups virtualization subcommittee.J. Carlos Pratt has worked in the digital storage industry since 1995, working mainly on static storage for the mining industry. After joining IBM in 1999, he worked on the channel adapter device driver for the virtua
39、l tape system. Later, Carlos joined the IBM disk test organization to help with the function and performance evaluation of the fiber channel connectivity between storage and different types of servers. In 2002, Carlos helped the IBM-Tivoli group validate the performance module of the multiple device
40、 manager, today known as Tivoli Productivity Center. Since 2006, Carlos has been a member of IBMs hardware systems storage perfor-mance organization. His current role in this organization is to provide architecture guidance for energy efficiency and high performance to the IBM storage portfolio. Car
41、los is member of the Storage Performance Council (SPC) and collaborated on the energy extensions for SPC-1 and SPC-2. Carlos is currently a member of the Storage Networking Industry Association (SNIA) Technical Council and is a co-chair of the Green Technical Work Group; he is also a member of the S
42、NIA Green Storage Initia-tive.Contributors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vAbout the Authors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viiForeword . . . . . . . . . . . . . . . . . . . . . . . . . .
43、. . . . . . . . . . . . . . . . . . . . . . xvPreface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xviiDisclaimers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xviiAcknowledgments. . . . . . . . . .
44、. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xixContributing Organizations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xixChapter 1Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.1 The Industrys NeedFilling the Void
45、. . . . . . . . . . . . . . . . . . . . . 11.2 Overview of this Book and Its Organization . . . . . . . . . . . . . . . . . 41.3 Summary of Benchmarks and Tools and Their Metrics . . . . . . . 11Chapter 2SPEC Power andPerformance Benchmark Methodology. . . . . . . . . . . . . . . . . . . . . . . . 1
46、52.1 Introduction to SPEC and the Methodology Document . . . . . . . 152.2 Document Overview: SPEC Power andPerformance Benchmark Methodology. . . . . . . . . . . . . . . . 15Chapter 3SPEC PTDaemon. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19Chapter 4SPEC Benchmarks . . . . . .
47、. . . . . . . . . . . . . . . . . . . . . . . . 214.1 SPECPower_ssj2008 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224.2 SPECvirt_sc2010 (Retired) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264.3 SPECvirt_sc2013 . . . . . . . . . . . . . . . . . . . . . . . . . .
48、. . . . . . . . . . 304.4 SPEComp2012. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 344.5 SPECweb2009 (Retired) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38Contentsxii ContentsChapter 5TPC-Energy. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
49、. . . . . . . .435.1 Introduction to the TPC and TPC-Energy . . . . . . . . . . . . . . . . . .435.2 TPC Benchmarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 445.3 Overview of the Energy Measuring System . . . . . . . . . . . . . . . .465.4 TPC-Energy Stages. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 465.5 TPC-Energy Execution and Validation . . . . . . . . . . . . . . . . . . . .535.6 TPC-Energy Tuning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .565.7 Sample Resul
