Network Storage Tools and Technologies for Storing Your Company’s Data James O’Reilly AMSTERDAM • BOSTON • HEIDELBERG • LONDON NEW YORK • OXFORD • PARIS • SAN DIEGO SAN FRANCISCO • SINGAPORE • SYDNEY • TOKYO Morgan Kaufmann is an imprint of Elsevier Morgan Kaufmann is an imprint of Elsevier 50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States Copyright © 2017 Elsevier Inc. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Details on how to seek permission, further information about the Publisher’s permissions policies and our arrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website: www.elsevier.com/permissions. This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may be noted herein). Notices Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understanding, changes in research methods, professional practices, or medical treatment may become necessary. Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any information, methods, compounds, or experiments described herein. In using such information or methods they should be mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility. To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in the material herein. Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the Library of Congress British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library ISBN: 978-0-12-803863-5 For information on all Morgan Kaufmann publications visit our website at https://www.elsevier.com/ Publisher: Joe Hayton Acquisition Editor: Brian Romer Editorial Project Manager: Amy Invernizzi Production Project Manager: Mohana Natarajan Designer: Victoria Pearson Typeset by TNQ Books and Journals Contents Acknowledgm.e.n.t. ....... .............. . .... Xl lntroduc.t..i..o..n. ......................... ...x .ii.i .. CHAPTE1R WhyS toraMgaet te...r...s... ............ .................. ............. 1 Referen.c.e.s. ..... ,""",'"" """,'. ,.. ...... ., ,',' , ... ,"""', ,"""" 4 ..... ............. CHAPTE2R StoraFgreo m3 0,00F0e e.t... ....................................... 7 WhatI sC omputeSrt orag..e..?.... ............................................. .... 7. StoraTgoed a.y, ............. .......8. ...... . TheL argCeo rporatl..O...t..l.. ....__. .............. ............................8 Mid-sized Operations ,9 TheS malBlu sine.s.s. ................................ .....9. ......... .. Storaigne3 Year...s.. .......................... .... ................1.0.. ....... TheD istantF utur2e0:1 9a nd Bey..o...n...d.. ................................ II Referen.c.e.s.. .......... . .. ..................................................1 .3 CHAPTE3R NetwoIrnkf rastrTuocdta.u.y.r.. e..... ...........................1 5 StoraAgree aN etworiknsT ransition................1..5.. ........ .... . HowS ANsW ork ... ..... 15 ....................... ............ ........ ............. iSCS(II nterSnmeatl l Computer SysteSmA Isn. ..t 1e8r face) SAN UseC ase.s.. ....................... .......................................... 19 UpsidoefsS ANs, ....................,..... ...... .. .2.0. , .... DownsidoefsS ANs. ....................................... .. ...2.0. ........ TheF uturoefS ANs. .. ..,.2.1. .... File(NrA S-Network-AtatchedS torag.e.). ...... .............. 2..2. . HowF ileWrosr k. , .., ..2.2. CompariNnAgS witShA N . . 25 .............. ........................ NFS, ClFS, L5iMnBu,x a,n dW indow..s, ,'"'' "2"7" """ FilUesre C ases"""," """""""" "". ,., """",',"2 8 """""""" Upsidoefsfi ler.s.. ..................... ........... .......2.8. ........... DownsidoefsF iJc.r.s. .........................................2.9. .............. FilMeyrt hs,,' "'"' '"' """., ., "",.,.,".,, ",, ,'",','"''3'"0'' ''' ' FilFeurt ur..e...s... ........................ .............. ..........................3 0 Scalingf ntfoi nity-ObSjteocrta. g.e ",,,,, 31 HowO bjecStt oraWgoer ks ",31 UseC asefso Orb jecStt orage """"",,33 UpsidteosO bjecStt orage ......................3.4. ................. DowilsitdoOe bsj ecStt orage.. ...........__ .... ....................3.4. ObjeScttor aeg Myths""""" """ "."."..",. ,,.,' " """,,,,,35 v vi Contents HardD rives............................. . ....................3..6 . ..... How HardD rvie sW ork.. ............................ ... .. ...3.7. ..... SAS andS ATA .................................................... ........3 9 UpsidetsoH ardD rives ......................4..0.. . ..... DownsidteosH ardD rives..." ,, ",.,', ....,.. .,..". ".,...4,.0". ,. ......... ., HardD rivMey ths ................................................... ....... 40 TheS olid-SRteavtoel ut.i...o..n... ....................... . .. .....4..2. . How DoF lasahn dS SDW ork? ...............4.3. ................. ]rnproviFnlga sCha paciyt. .......................... . . .......... 45 DeletiPorno blemisnF l ash. .. .......................4 ..5.. ............ PackagiFnlga sh............. ..............4.5. ................ TheN VMe Revolut.i.o.n. ................. ......4..6.. ........ ..... Upsidoefs Fl.a...s...h... ....................................................... .46 DownsidoefsF las...h.. ................................ . .. ..4.7. . Mythosf F lash andS SD ................................47 . . TheF utuorfeF lasahn dS SD.. .........................................4..9. RDMA-TheI mpacto fA llF·la sh andH ybriAdr ra.y...s. ...........5 0 How RDMAW orks................ ...... .................5..0 . ...... UseC asefso rR DMA .... .... ...........5..1 Compariosn ofInfiniiBWaAnRdP.,a ndR oCE .........5 2 UpsideosfR DMA....... .... .....................5..2 . ...... DownsidoefsR DM A ....................................5.2. ..................... TheF uturoef RDMA.... .............................. ... . .5.3. . OprimizjtnhgeD atacen.t..e....r.. ........................................ ........5 3 SonleU seC ase.s.. ................... .. ..............S..4 . ........... Q&A ............................ ...................................... .............. ........5 6 lsth eA pproacthoS rorage Differeinntt h eH PC/Big DataW orld?. ........... ................................... ........5..6.. . .......... How Do FalliSnogl id-SPtraitceA efsf ecPtl anni..n..g..?. . .56. . WhatA bouHty bridA rrays?. ................................5...6... ...... .. ShoulIdC onsidEetrh emeStA Ns?. .......................5..7. ......... Referen.c..e..s.. ................... .. ................5..7 . ...... CHAPTE4R StoraSgoeft war...e... .................................................. 59 TraditiSoonluatilo ns .......... .... ....................6..0 . ...... Pools, LaankdeO sc,e an...s. .......... .......... .. ..............................6 1 CompresDseidounp,l icanadtA lilTo hna,t ....................... 62 Deduplica.t.i.o.n. . ....... ............. ................ ... 62 COlnpress.i.o.n. ............. . .......................6..2 . ...... ProblWeimthsC o mpressainodnD eduplicaTtoidoa.n.y.. ...6 3 Contentsv i i Open-SouSrocfet ware .....6.3. ..... OpentSa ck ............... ... .. ........6.4. ............ ..... . ..... Ceph 69 .... ..................................................... .......................... Gluste.r.F.S .. ..7.4. ... VirtualizTaotoil"os"n ', """',...,.".". ...,. ......... .,.".,...,.".'.". .". .. , 75 ., " VMware ............ ... .. .........7 .5 ......... ...... . . ....... . ...... ...... ....... ...... . VirtuSaAl s ................. .................. ......7.6. .............. TheO peraliSnygs teSml OraSgtea c...k... ............. ...... ..7.7. .... TheT raditiSoCnSaISl t ack .............. .......................... 77 TheN VMe Stacakn dB eyon.d.. ..................... .. .. ...... 78 EthernDerti vaensd K eylDAactcae Psrosto cols. ..........8.1. ..... References ... ... . ... .. 82 ... .......... . ........... ......... .............. .......... ........... CHAPTE5R Software-DSetfoirnae..g.d.e... .........................." . ...8.3. .. Prelu.d.e. ............ .....................8. 3.. .. ........ Software-DeSfilnOerda ge . 84 ............................. .. .......................... A CompleteelwyA pproach. ... .....8..5. . WhoA ret heP layers? .....8.7. .... "LegoS"t ora.g.e.... .............................................. .. .....................8 7 ConnectitnhgeP iec.e...s. ........................................................8.8. . UnifieSdt oraAgpep lian.c..e...s.. ................................................. 88 AgiliatnydF lexibility..................... ....8..9.. ......... ..... The[ mplicatoifSo DnSst ot heD atacente.r... ......................8 9 SDSa ndt he Cl..o..u...d.. ............................................................. 90 TheC urrSetnatto efS DS. .............................................. .........9..0. TheF utuorfet heS toraIgned us.t.r.y. ........ ...... ......9..1.. .......... . References . . . . 93 .. .............. ...................... .............. ............ ............. CHAPTE6R TodayH'ostI ssu.e...s... ...............................................9 5 NAS VersuSsA VersuObsj ecStt orage ........................9 5 Etherneatn dt heEn do ft heS AN ........................9.7. ............... Commoditiaznadt thiSeto onr aNgeet wor...k:... ..... .. 1.0.0. .. Referenc.e.s.. ...... ..................... .......... ..... .......... ...1.0.3. ... ... CHAPTE7R Tunitnhg eN etwo..r..k.... ............................................ 105 Gettiunpg1 0S pee.d... ........................ ................ .....1 .0.5. ... LetM e ReachO m toT hosPee oplNeo w'. .. ....1.0.5. ... Tuninfgo rF asSty stems. .."........ .... .... .". ....."."..... 1.0.6. ....... The( VirtuDaelsk)t op...................... ............. .. ..1 07 TheI nsranScteo r.e. ................................ ...... ..1.0.8. ... . NetworkSetdo rageA-NS ...." ................ .... ... ..1 .0.9. NetworkSetdo rage-NFAiSl e.r...... ....... ........ ....1..1I.. .. viii Contents TuninSgS D .................. .. .............1. .1.2. ..... The Vinual SA.N.. ................ ........ ......... . 1.1.5 Tuninfgo rV irltalCll uste..r...s.. ............................... __ .......1 15 Tuninfgo rB igD at.a.. ......... ,.," .............. 1.J7. .............. Tunintgh Se lOraNgeet wokr. ......" "". ...............".1. .1..7. ........... Tuninfgo rt heC IOL.I...d.. ............................. ......................1 18 TheN ew StoraTgiee rs......_... ............................. .. ..1.1.9. .. CachinDga ta.. .... .................... ...................1 .2.0. ...... SSD and thFei lSey stem. ...............................1...2...1... .. ............... \Vhat Tuning MCeoaunli dn 5 -YeaTrism'e .. . ........1. 2.3. ........ Referen.c..e.. s.... .................. ... ..................1. .2.4. ...... CHAPTE8R BigD at..a.. .................................................................. 127 AddressiBnigg -DatBao ttlene,c ks ....".".. ......1. .27. .... ", NetwokrB andwihd. t... .... ............................1. 2.7. ........... SpeeidnIgn -MemorCyo mputin....g.. ........................ .......1 28 Key/ValSLtioer a.g.e. .. ..... .. .................1.. .2..8. ..... BigD ataS tora.g.e. ....... .......... ...................... .1.2.9. .... ServeSrto rage VersuNse tworSkt orage .................... .1 34 TechnoloCghya ngintgh eP ictu.r.e. ..... ......... 137 Bandwidatnhd S cal.e.. ........ ........................ .....................1 38 FjlSey stemfso Bri gD ata. ... ........ ................ .1.4.0. ..... References. .................................................... ....................1 48 CHAPTE9R High-PerforCmoamnpcuet i...n...g.... .........................1 51 MajoSrc ientEixfpiecr ime.n..t...s.. ..................... .................... .1 52 BigS imulati..o...n...s... .. .................................................... ......1 53 SurveiJlaSnycset em..s.. ................ . .................... S.14. ....... TheT radiFnlgo o.r.. ...................................................................I .SS High-PerformCaonmcpeu tinClgo usd. ............ .. . ....1 .5 6 VideEod ilj..n..g.. ............................................................... ......1 57 Oila ndG a.s................... .................. 1..5..8.. ................ . LatencIysK ING'. .................. ........... ..1..5..8.. ... ........ ConvergiRnegm otDei reMcetm oryA ccess. .........1 .6.0. ........ Referen.c..e.. s.... .................. ... ..................1. .6.1. ...... CHAPTE1R0 TheC lou..d.... ............................................................. 163 WhatI st heC lou?d. ....................... ...........................................1. 65 CloudH ardware........................... ........... ......... .1 6.7. ...... TheF uturoefC louHda rdw..e a . ........................................ .......1 69 ClouSdo ftw.a.r..e. ........... ........................ ...........1..7.0. ....... OpenSlaScokf tw.. ea. ............ .... ............1 .7.1. ...... Contents ix AWS ...... .... ....1.7.3. .. MicrosAofztu r.e.. ......... ................ . .....................1 75 Googel .....................................................1..7..6.. ................... VMware....................... ............... . .1.7.7. .. TheR est..... .........................1.7.8. .............................................. TheC hanginDga tacenDteenrs i, tCyoolinagn,dP ower. ....1.7.9. ... UsinCgl ouSdt ora.g.e. ...............................................18.2. ............ HybriCdl ouds. . .............. ....1..8..3.. ......... ........ EvolviCnlgo ud.s.. ...................... .............. .......................1. 85 Hardwaroeft hCel ouOvde rt heN extD ecade ..... ...... .... 185 Softwaorfet heC louOdv ert heN extD ecad.e. ... .... ..1.8.7. .... NetworkifnogrC loud Stora.g...e.. ............ .......................1. 89 Backuapn dA rchiving ......................................................1.9 2 HybriCdl oudasn dD ata Governanc..e... ............................ 193 Contajnaenrds T heiIrm pac0t11 Storage Net.w.o.r.k..s.. . .1.95 Wiltlh eP rivaDtaet acenStuerrv iv.e...?.. .......................... 1.97 Referen.c...e..s .. .......................................1.9.9. ... CHAPTE1R1 DatIan teg..r..i.t.....y.... ..............". . .............2.0.3. .............. RAIDa ndI tPsr oble.m.s.. ........................................................ 205 How RAIDW orks.?.. .................. ................... ....................2 05 RAID Myths .............................. .. ..2.0.6. ... Replica.t...i.o...n.. .................................................. ....................2 07 ErasuCroed ing.. ................................ ......................................2. 08 DisastPerro tectiinto hne C loud... .......................................... 2.,0 9 Referen.c.e.s. . ... .. ...2.1.1. .. CHAPTE1R2 DatSae cur.i...t.y. ..........". . ..........".. ................2..1..3 . ....... LosinYgou rD at.a. .................................................................2..·1..4. Dataat R es.t. ............................................................2.1.5. ...... Datai nT rans.i.t... ....................... ......................................2. 18 ColdD ata............. . .......... ......... ....... ..2..1..8.. . . Protect:Cilnogu Dda ta. ............... ....... ........2..2..0.. ......... ..... Software-as-a-(SSearavSi)c .e. ....... ....... .2..2..1... . HybriCdl oudasn dD atGao vern.a...n..c....e.. ........ .. .. ...2.2.2. .. TheM odern GovernanAcpep roac...h... .......... ...... .............2 22 Encryp(io.n. ......................................................2..2..3.. ........... .. ImportaEnntc ryptSicohne me.s. ..................2.2..3. . Commanda ndS tatMuess sag.e.s. ........ .... .....2..2..4.. ...... . StandaFridlF eo rma.t.s. .................... ........ .......2..2..5.. .. ... KeyM anagement.. ....2.2.5. ... x Contents informatRiiognh tMsa nagement. ... ..............................2 26 References. ........................ ............ 2..2..7. ............ CHAPTE1R3 Otnh eH oriz.o....n.. ....................................................2 31 Solid-RSetpaalcteSesp inning Rust. ......................... .. .231 TheI mpacotn A pplianDcees ig..n... ...........__ .............. ......2 34 CollateCroanls iderations .................2..3.5.. ........... . NVDIMMs:C hangintgh Bea lancoefS tora.g..e. .....................2. 36 TheH ybridM emoryC ube.. ....................................... ...239 VirtuSaAlN s_ _ ............_ _ ..................................._._ ............ __ ......2 40 InternoefTt h ing.s. .............. . ......................2 .4.1. ..... Retiaal nd1 01' ....................................__ ............_ _ ......2 42 TheO theBri gD ata. .................................2....4...3.. ................ Referen,c. e..s . ...... ... ......... ............2.4.3. ................ , " , , " .." " ,., CHAPTE1R4 J ust OtvheeHr o riz...o...n... ........................................ 245 NAND Kille...r...s.. ............. ....................................... .............. . 245 Phase-Change Memory ..............2. 4.6. ................ Resj"iRvAeM .................................................... _ ......2 47 SpintronMiRcAsM ......... ...........................__ ..........._.__ .....2 48 Graphen..e... ........... ................ ..................2 .4.9. ....... FurtheOru t. ......2.5.0. ..... References .... ..................2 .5.0. ......................... Conclusi...o...n... ................................ ... ...................2 .5.3. ...... A BrieHfi stoorfyS torageN etworki.n...g.... ...................................................... 255 Glossar...y... ..........................._.__ .........._ ___ ..................................._._ .......... ____ ......2 57 Ind.e.x. ............... .............2. 6.]. .............. , Acknowledgment In my own journey through the magic world of storage, I have been lucky to have the support of some great mentors. I would like to recognize four in particular who had a major impact on my carrier, my knowledge base, and my achievements. In my first job at International Computers in Stevenage in the United Kingdom, Jim Gross allowed me the freedom to run some big projects, saved me from implod- ing, and thoroughly grounded me in the how-to of doing good designs. When I moved on to Burroughs, Arnie Spielberg gave me a lot of sage advice, while Mark Lutvak acted as my marketing guru, and was instrumental in getting me to move to the US. Arnie, incidentally, was Steven Spielberg’s dad and I still remem- ber the day that he came up to me and said, “My boy Steven, he’s decided to go into the movie business. I’m worried for him…the business is full of gonifs. Still, I’ve got to stand back and let him try!”. Ray Valle proved to be a good friend while I was at Memorex and beyond. When I went on to NCR, Dan Pigott, the GM, allowed me a great deal of room to innovate, all the while protecting me from the intense politics of the company. Dan later hired me again to run the PC Division of Memorex-Telex. There have been many others over the years. The NCR team in Wichita, Kansas, was a joy to work with. They learned to enjoy innovation and not fear the challenges. I would go so far as to say they were the best engineering team in the IT industry at the time, though of course I might not be totally objective. I would like to thank my editors at Elsevier, without whom this book would never have seen the light. Brian Romer approached me to write a book on the subject and guided me through the early stages while selling Elsevier on the idea. Amy Invernizzi picked up the herding task to get me to complete close to schedule, and she mentored me throughout the writing process. They are both great people to work with and made a daunting task somehow manageable. Obviously, I have missed a lot of names. Many people have worked with me over the years and most I remember with respect and fondness. Their willingnesses to help projects move forward quickly and to put in more-than-the-required effort have allowed some great team successes and I will always remain grateful. Finally, I would like to thank my wife Angela for uncomplaining support over these many years. We have moved house as I changed jobs and made new sets of friends too many times to count. Through it all, she has supported me, put up with my ranting and periods of intense concentration on the job, laughed at my silliness and listened to my ideas. She also knows when to feed me tea and her glorious home- made mince pies! Without her, I don’t think I could have done this. xi