Ken Cantrell / NetApp Mark Rogov / EMC July 30, 2015.

Slides:



Advertisements
Similar presentations
Circling Back To Littles Law Now that we have tools to gather information.
Advertisements

Calderdale Children & Young Peoples Service
Tag line, tag line Perforce Benchmark with PAM over NFS, FCP & iSCSI Bikash R. Choudhury.
STANFORD UNIVERSITY INFORMATION TECHNOLOGY SERVICES IT Services Storage And Backup Low Cost Central Storage (LCCS) January 9,
An Adaptable Benchmark for MPFS Performance Testing A Master Thesis Presentation Yubing Wang Advisor: Prof. Mark Claypool.
The Impacts of SSD in Your Storage System Author: David Dale, SNIA (NetApp)
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.
Nexenta Proprietary Global Leader in Software Defined Storage Nexenta Technical Sales Professional (NTSP) COURSE CONTENT.
® IBM Software Group © 2012 IBM Corporation OPTIM Data Studio – Jon Sayles, IBM/Rational November, 2012.
File System Benchmarking
Copyright DataDirect Networks - All Rights Reserved - Not reproducible without express written permission Adventures Installing Infiniband Storage Randy.
Copyright © 2000,2003 Storage Networking Industry Association “Iconic images” from the SNIA shared storage model These slides are a subset of the images.
Storage Benchmark Proposal Edgar StPierre, EMC. Proposal Project Name: STORPERF Repo Name: STORPERF Category: Requirements Project Lead: Edgar StPierre,
HEAnet Centralised NAS Storage Justin Hourigan, Senior Network Engineer, HEAnet Limited.
MESSAGING IN THE MEDIA TOP 10 DRUGS BASED ON CREDIBILITY INDEX IN THE LITERATURE JOHN KNAPP, DIRECTOR OF ANALYTICS MAY 2009 Your use of these materials.
PRESENTATION TITLE GOES HERE Presenter Name/Company.
RD Connection Brokers Personal Desktop Pooled Desktops RD WEB Session Hosts VDI Corp LAN User login Get list of published apps & collections User.
Therapeutic area breakdown of lead projects in development for first launch in 2005 APRIL 2007 SOURCE: CMR INTERNATIONAL PERFORMANCE METRICS PROGRAM ©
IBIS-AMI and Direction Decisions
Data in the Cloud – I Parallel Databases The Google File System Parallel File Systems.
WSTIP Assessing Economic Funding Needs Strictly Private and Confidential Draft Result Update June 28, 2012 DISCUSSION PURPOSES ONLY.
Proportion of biotech-derived active substances in development for first launch 31st December APRIL 2007 SOURCE: CMR INTERNATIONAL PERFORMANCE.
1 CS/COE0447 Computer Organization & Assembly Language CHAPTER 4 Assessing and Understanding Performance.
Session Objectives And Takeaways A word on Perf & VDI architecture.
ONStor Pantera 3110 ONStor NAS. Copyright 2008 · ONStor Confidential Pantera 3110 – An Integrated Channel only NAS  Integrated standalone NAS system.
August 2008 WD VelociRaptor 10K RPM 2.5” and 3.5” Enterprise SATA.
Legal Disclaimers Accuracy Every effort is made to provide information that is accurate. However any information contained in this website or the “article.
Held in partnership with Creating Connections. Held in partnership with Creating Connections This document has been prepared for general guidance on matters.
Proportion of total R&D expenditure by therapeutic area in 2005 APRIL 2007 SOURCE: CMR INTERNATIONAL PERFORMANCE METRICS PROGRAM © THOMSON REUTERS Your.
Global R&D expenditure, development times, global pharmaceutical sales and new molecular entity output MARCH 2008 SOURCE: CMR INTERNATIONAL.
2005/2006 Expenditure on biotech derived entities had remained constant APRIL 2007 SOURCE: CMR INTERNATIONAL PERFORMANCE METRICS PROGRAM © THOMSON REUTERS.
Number of pharmaceutical patent applications and granted patents for 5 major patent issuing authorities JUNE 2008 SOURCE: DERWENT WORLD PATENTS.
Ken Cantrell / NetApp Mark Rogov / EMC J Metz / Cisco October 21, 2015.
PRESENTATION TITLE GOES HERE Emerald NAS Extensions Chuck Paridon Performance Architect H-P Enterprise Data Contributed by Nick Principe – EMC, Demartek.
Regulatory approval times between submission and marketing authorization approval APRIL 2007 SOURCE: CMR INTERNATIONAL PERFORMANCE METRICS PROGRAM.
Jérôme Jaussaud, Senior Product Manager
Allocation of R&D expenditure in 2005 between chemical entity and biological or biotech active substances APRIL 2007 SOURCE: CMR INTERNATIONAL PERFORMANCE.
Ken Cantrell, NetApp Mark Rogov, EMC David Fair, SNIA ESF Chair, Intel March 8, 2016.
Give a quick overview of the Microsoft Dynamics AX 2012 training system architecture Discuss ‘Day in the life’ benchmark study Provide the first sizing.
This courseware is copyrighted © 2016 gtslearning. No part of this courseware or any training material supplied by gtslearning International Limited to.
IBM Innovate 2012 Title Presenter’s Name Presenter’s Title, Organization Presenter’s Address Session Track Number (if applicable)
HUAWEI TECHNOLOGIES CO., LTD. Huawei Storage ISM Management Pre-sales Product Training Materials Easy and Efficient WEU IT Solution Team.
1 / ? YYYY.MM.DD / 홍길동 책임 / xxxxxx 팀 Understanding Why Demand for SSDs is Soaring Ryan Smith Sr. Manager – SSD Product Marketing, Samsung Semiconductor,
RaboDirect Financial Health Barometer 2016
Storage Performance Benchmarking:
SharePoint Solutions Architect, Protiviti
Bits and bytes September 19, 2017.
Credit Risk Management Mark Schulz Lugano, 22 September 2016
BIOLOGY / CHEMISTRY CITATIONS GROWTH
Microsoft Ignite NZ October 2016 SKYCITY, Auckland
VNX Storage Report Project: Sample VNX Report Project ID:
XtremIO Storage Array Asset Profile
Unity Storage Array Profile
Messaging in the Media TOP 10 AUDIENCES BY THERAPY AREA
MDC-B345 Hyper-V Storage Mathew John. MDC-B345 Hyper-V Storage Mathew John.
2006 Revenue continued to depend on a relatively small number of products APRIL 2007 SOURCE: CMR INTERNATIONAL PERFORMANCE METRICS PROGRAM © THOMSON REUTERS.
, R&D expenditure showed 6% growth annually
Building a Database on S3
2002–2006 Mid and Other sized companies continued to launch the majority of NMEs APRIL 2007 SOURCE: CMR INTERNATIONAL PERFORMANCE METRICS PROGRAM © THOMSON.
IBM Blockchain An Enterprise Deployment of a Distributed Consensus-based Transaction Log Ben Smith & Kostantinos Christidis 1 ©2016 IBM Corporation.
Assessment Findings System Professional <Insert Consultant Name>
12/3/2018 Desktop Virtualization Corey Hynes Kyle Rosenthal President Technical Lead HynesITe Inc Spider Consulting @windowspcguy.
Proportion of total R&D expenditure on alliances or joint ventures by stage of R&D in 2005 APRIL 2007 SOURCE: CMR INTERNATIONAL PERFORMANCE METRICS PROGRAM.
% of the projects in development in the leading therapeutic areas were considered to have a novel mode of action APRIL 2007 SOURCE: CMR INTERNATIONAL.
Environmental Upgrade Finance
Pipeline size of Major companies had grown at a steadier rate than that of smaller companies APRIL 2007 SOURCE: CMR INTERNATIONAL PERFORMANCE.
Development time for new molecular entities first launched onto the world market between MARCH 2008 SOURCE: CMR INTERNATIONAL PERFORMANCE METRICS.
Cycle times decreased in larger but not smaller, companies
Proportion of total R&D expenditure by function in 2005
Analysis on IEEE n MAC Efficiency
Presentation transcript:

Ken Cantrell / NetApp Mark Rogov / EMC July 30, 2015

© 2015 Storage Networking Industry Association. All Rights Reserved. About The Speakers 2 Mark Rogov EMC Systems Dr. J Metz Cisco R&D Ken Cantrell NetApp Manager Perf

© 2015 Storage Networking Industry Association. All Rights Reserved. SNIA Legal Notice The material contained in this tutorial is copyrighted by the SNIA unless otherwise noted. Member companies and individual members may use this material in presentations and literature under the following conditions: Any slide or slides used must be reproduced in their entirety without modification The SNIA must be acknowledged as the source of any material used in the body of any document containing material from these presentations. This presentation is a project of the SNIA Education Committee. Neither the author nor the presenter is an attorney and nothing in this presentation is intended to be, or should be construed as legal advice or an opinion of counsel. If you need legal advice or a legal opinion please contact your attorney. The information presented herein represents the author's personal opinion and current understanding of the relevant issues involved. The author, the presenter, and the SNIA do not assume any responsibility or liability for damages arising out of any reliance on or use of this information. NO WARRANTIES, EXPRESS OR IMPLIED. USE AT YOUR OWN RISK. 3

© 2015 Storage Networking Industry Association. All Rights Reserved. Storage Performance Benchmarking 4 SYSTEM UNDER TEST METRICS AND TERMINOLOGY FILE COMPONENTS BLOCK COMPONENTS WORKLOAD DEFINITIONS

© 2015 Storage Networking Industry Association. All Rights Reserved. Storage Performance Benchmarking 5 SYSTEM UNDER TEST METRICS AND TERMINOLOGY FILE COMPONENTS BLOCK COMPONENTS WORKLOAD DEFINITIONS TODAY FUTURE WEBCASTS INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. BUSINESS RUN PRODUCTION APPLICATIONS 6 Three Great Buckets Of Testing DATABASE COMPUTE FARM VDI INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. BUSINESS RUN PRODUCTION APPLICATIONS BAKE-OFFS SINGLE OBSCURE METRIC 7 Three Great Buckets Of Testing UN*X DD IOMETER IOZONE VDBENCH DATABASE COMPUTE FARM VDI INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. BUSINESS RUN PRODUCTION APPLICATIONS BAKE-OFFS SINGLE OBSCURE METRIC 8 Three Great Buckets Of Testing UN*X DD IOMETER IOZONE VDBENCH DATABASE COMPUTE FARM VDI INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. BENCHMARKS BALANCED PROFESSIONAL APPROACH BUSINESS RUN PRODUCTION APPLICATIONS BAKE-OFFS SINGLE OBSCURE METRIC 9 Three Great Buckets Of Testing UN*X DD IOMETER IOZONE VDBENCH DATABASE COMPUTE FARM VDI LOGIN VSI VMMARK SPEC SFS SPC1 INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. 10 Block / SAN Terminology IO BLOCK / SAN PROTOCOL FC/iSCSI/FCoE Target Initiator INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. 11 Block / SAN Terminology IO Target BLOCK / SAN PROTOCOL FC/iSCSI/FCoE IO OPERATION AKA TYPE: READ WRITE Initiator INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. Block / SAN Operations 12 Source: Graphic: INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. 13 Block / SAN Terminology IO Target BLOCK / SAN PROTOCOL FC/iSCSI/FCoE IO OPERATION AKA TYPE: READ WRITE Initiator IOPS COUNT EVERY IO OPERATION PER SECOND THROUGHPUT SAME AS IOPS INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. Where To Measure IOPS? 14 HOSTS S TORAGE CONTROLLER D ISK DRIVES INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. Where To Measure IOPS? 15 HOSTS S TORAGE CONTROLLER D ISK DRIVES F RONT E ND 1500 INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. Where To Measure IOPS? 16 HOSTS S TORAGE CONTROLLER D ISK DRIVES IDLE F RONT E ND B ACK E ND INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. 17 Block / SAN Terminology IO Target BLOCK / SAN PROTOCOL FC/iSCSI/FCoE IO OPERATION AKA TYPE: READ WRITE Initiator IOPS COUNT EVERY IO OPERATION PER SECOND MB/S PAYLOAD SUM OF EVERY OPERATION PER SECOND BANDWIDTH SAME AS MB/S INTROBLOCKFILEGRAPH THROUGHPUT SAME AS IOPS

© 2015 Storage Networking Industry Association. All Rights Reserved TERABYTES TB GIGABYTES GB MB/s Vs. MiB/s 18 MEGABYTES MB KILOBYTES KB TEBIBYTES TiB GIBIBYTES GiB MEBIBYTES MiB KIBIBYTES KiB 10% 7% 5% 2% DECIMALBINARY INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. 19 Block / SAN Terminology IO Target BLOCK / SAN PROTOCOL FC/iSCSI/FCoE IO OPERATION AKA TYPE: READ WRITE IO SIZE SIZE OF DATA PAYLOAD IN AN IO OPERATION Initiator IOPS COUNT EVERY IO OPERATION PER SECOND MB/S PAYLOAD SUM OF EVERY OPERATION PER SECOND BANDWIDTH SAME AS MB/S INTROBLOCKFILEGRAPH THROUGHPUT SAME AS IOPS

© 2015 Storage Networking Industry Association. All Rights Reserved. MB/s (or MiB/s), IO Size, And IOPS 20 KEEPING IOPS CONSTANT, L ARGER I/O SIZES PRODUCE GREATER MiB/ S 512B 1KiB 4KiB INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. MB/s (or MiB/s), IO Size, And IOPS 21 AS IO S IZE INCREASES IOPS DECREASE FOR THE SAME MiB/ S AS IO S IZE INCREASES IOPS DECREASE FOR THE SAME MiB/ S 16KiB8KiB4KiB INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. IO SIZE SIZE OF DATA PAYLOAD IN AN IO OPERATION 22 Block / SAN Terminology IO Target BLOCK / SAN PROTOCOL FC/iSCSI/FCoE IO OPERATION AKA TYPE: READ WRITE Initiator IOPS COUNT EVERY IO OPERATION PER SECOND MB/S PAYLOAD SUM OF EVERY OPERATION PER SECOND BANDWIDTH SAME AS MB/S RESPONSE TIME TIME TARGET TAKES TO REPLY TO AN IO LATENCY SAME AS RESPONSE TIME INTROBLOCKFILEGRAPH THROUGHPUT SAME AS IOPS

© 2015 Storage Networking Industry Association. All Rights Reserved. Aggregate Property Of Response Time 23 D ISK DRIVES S TORAGE CONTROLLER HOSTS INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. Aggregate Property Of Response Time 24 D ISK DRIVES S TORAGE CONTROLLER HOSTS  WRITE ≥ 1ms 1ms INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. Aggregate Property Of Response Time 25 D ISK DRIVES S TORAGE CONTROLLER HOSTS   READ  WRITE ≥ 1.5ms ≥ 1ms 1ms 0.5ms INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. Aggregate Property Of Response Time 26 D ISK DRIVES S TORAGE CONTROLLER HOSTS   READ  WRITE ≥ 1.5ms ≥ 1ms READ  1ms ≥ 5ms 4ms 0.5ms INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. 27 File / NAS Terminology OP FILE PROTOCOL NFS/SMB/FTP Client Server INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. File / NAS Terminology 28 OP FILE PROTOCOL NFS/SMB/FTP Client Server FILE OPERATION SEVERAL DOZEN DEPENDING ON PROTOCOL INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. NFSv4 Operations 29 Source: Graphic: INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. SMB3 Operations 30 Source: (use [MS-SMB2].pdf) Graphic: INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. File / NAS Terminology 31 OP FILE PROTOCOL NFS/SMB/FTP Client Server FILE OPERATION SEVERAL DOZEN DEPENDING ON PROTOCOL OPS COUNT EVERY PROTOCOL OPERATION PER SECOND INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. File / NAS Terminology 32 OP FILE PROTOCOL NFS/SMB/FTP Client Server FILE OPERATION SEVERAL DOZEN DEPENDING ON PROTOCOL OPS COUNT EVERY PROTOCOL OPERATION PER SECOND MB/S PAYLOAD SUM OF EVERY OPERATION PER SECOND THROUGHPUT SAME AS MB/S BANDWIDTH SAME AS MB/S INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. File / NAS Terminology 33 OP FILE PROTOCOL NFS/SMB/FTP Client Server FILE OPERATION SEVERAL DOZEN DEPENDING ON PROTOCOL OPS COUNT EVERY PROTOCOL OPERATION PER SECOND MB/S PAYLOAD SUM OF EVERY OPERATION PER SECOND THROUGHPUT SAME AS MB/S BANDWIDTH SAME AS MB/S INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. File / NAS Terminology 34 OP FILE PROTOCOL NFS/SMB/FTP Client Server FILE OPERATION SEVERAL DOZEN DEPENDING ON PROTOCOL OPS COUNT EVERY PROTOCOL OPERATION PER SECOND MB/S PAYLOAD SUM OF EVERY OPERATION PER SECOND THROUGHPUT SAME AS MB/S BANDWIDTH SAME AS MB/S RESPONSE TIME TIME TARGET TAKES TO REPLY TO AN IO LATENCY SAME AS RESPONSE TIME INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. Difference Between OPS and MB/s (MiB/s) Two Otherwise Identical Runs With Different NFS Operations Mix (64 KiB IO Size) 35 INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. Difference Between OPS and MB/s (MiB/s) Two Otherwise Identical Runs With Different NFS Operations Mix (64 KiB IO Size) 36 High MetaData Operations High Data Operations INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. Difference Between OPS and MB/s (MiB/s) Two Otherwise Identical Runs With Different NFS Operations Mix (64 KiB IO Size) 37 High MetaData Operations High Data Operations INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. Graph Fun Or Why “It Depends” Is Relevant? 38 5,000 10,000 INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. Graph Fun Or Why “It Depends” Is Relevant? 39 INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. Graph Fun Or Why “It Depends” Is Relevant? 40 Operations per sec, OPS Response time, ms INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. Graph Fun Or Why “It Depends” Is Relevant? 41 Operations per sec, OPS Response time, ms INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. M AXIMUM PERFORMANCE NEED M AXIMUM PERFORMANCE NEED Graph Fun Or Why “It Depends” Is Relevant? 42 Operations per sec, OPS Response time, ms INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. M INIMUM PERFORMANCE REQUIREMENT M INIMUM PERFORMANCE REQUIREMENT Graph Fun Or Why “It Depends” Is Relevant? 43 Operations per sec, OPS Response time, ms INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. A CCEPTABLE L ATENCY B AND Graph Fun Or Why “It Depends” Is Relevant? 44 Operations per sec, OPS Response time, ms INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. A CCEPTABLE L ATENCY B AND Variance 45 Operations per sec, OPS Response time, ms INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. Other Metrics To Consider 46 Operations per sec, OPS Response time, ms $2/OP $10/OP $1.1/OP $5.5/OP INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. Other Metrics To Consider 47 Operations per sec, OPS Response time, ms $1.1/OP $5.5/OP $25K/U $1K/U $40/W $140/W 17 ms 10 ms INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. Other Metrics To Consider 48 Operations per sec, OPS Response time, ms $1.1/OP $5.5/OP $25K/U $1K/U $40/W $140/W 17 ms 10 ms CONTEXT MAKES METRICS MATTER CONTEXT MAKES METRICS MATTER INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. Summary 49 GRAPH FUN CONTEXT MAKES METRICS MATTER CONTEXT MAKES METRICS MATTER OPS COUNT EVERY PROTOCOL OPERATION PER SECOND MB/S PAYLOAD SUM OF EVERY OPERATION PER SECOND TERMINOLOGY IOPS COUNT EVERY IO OPERATION PER SECOND RESPONSE TIME TIME TARGET TAKES TO REPLY TO AN IO INTROBLOCKFILEGRAPH

© 2015 Storage Networking Industry Association. All Rights Reserved. Storage Performance Benchmarking 50 SYSTEM UNDER TEST METRICS AND TERMINOLOGY FILE COMPONENTS BLOCK COMPONENTS WORKLOAD DEFINITIONS TODAY FUTURE WEBCASTS

© 2015 Storage Networking Industry Association. All Rights Reserved. After This Webcast 51 This Webcast and a PDF of the slides will be posted to the SNIA Ethernet Storage Forum (ESF) website and available on-demand A full Q&A from this webcast, including answers to questions we couldn't get to today, will be posted to the SNIA-ESF blog Follow us on Next Webcast - October 21 st “Storage Performance Benchmarking: Part 2”