Characterizing Data-Intensive Workloads on Modern Disk Arrays

Size: px
Start display at page:

Download "Characterizing Data-Intensive Workloads on Modern Disk Arrays"

Transcription

1 Characterizing Data-Intensive Workloads on Modern Disk Arrays Guillermo Alvarez, Kimberly Keeton, Erik Riedel, and Mustafa Uysal Labs Storage Systems Program 4 th Workshop on Computer Architecture Evaluation using Commercial Workloads January 21, 2001 CAECW-01, 0 Motivation New features for midrange and high-end disk arrays have significant impact on performance: Large caches (256MBx2 in HP FC-60, 64GB in HP XP512) LUN-level prefetching Stream detection (for prefetching and caching) Degraded operation and on-line reconstruction Our understanding of how these array features perform for complex workloads is insufficient Most workload characterizations don t have attributes to capture the effectiveness of these features Goals: Quantify performance effects of array features Identify workload behaviors that leverage these features CAECW-01, 1 1

2 Potential benefits Understanding how a real array works Quantifying the (relative) influence of each array feature on overall performance Array architecture: Which configuration knobs and monitoring points are the right ones? Improving the fidelity of our workload specifications Better recreation of workloads (see Kurmas, et al.) Better prediction of workload performance Laying the groundwork for more realistic array models Behavioral models require detailed understanding Most work has focused on very simple (stochastic) workloads CAECW-01, 2 Outline Motivation Description of our approach Experimental infrastructure Prefetching investigation Conclusions and ongoing/future work CAECW-01, 3 2

3 Approach: find and quantify sweet spots Identify array features ( sweet spots ) that impact performance Identify workload behaviors for which the features may work especially well/badly Run synthetic experiments that exhibit one such behavior at a time Establish limitations of each sweet spot Quantify performance (latency) impact of each sweet spot Propose workload metrics that capture each behavior Replay traces from real applications Measure new workload metrics on each trace Explain trace s performance as a function of metric CAECW-01, 4 Experimental infrastructure: HP FC-60 disk array Characteristic Cache size Disk enclosures Disks per enclosure Disk capacity Total capacity Our configuration 256 MB x GB 0.5 TB (raw) Max configuration 512 MB x GB 4.4 TB (raw) Two controllers One fibre channel port/controller Each connected to all enclosures Six ultra-wide SCSI interfaces to controllers (40 MB/s each) Cache: distributed across controllers Write-back policy, with writes mirrored across controllers Top observed performance per controller: 84 MB/s, 5400 IO/s CAECW-01, 5 3

4 Identify array sweet spots Prefetching Behaviors How much useless read-ahead work? Holes in sequential runs? Interference between streams? Array control knobs Minimum prefetch Prefetch multiplier Disable length Caching Burstiness Stream detection Degraded mode CAECW-01, 6 Identify workload behaviors Issue: Prefetching may still be effective in face of non-strictly sequential workload behavior Additional workload behaviors to consider: Jumps/holes in sequential runs Jump distance = gap between next address and expected next address if accesses are sequential Interleaved accesses from different streams Interference distance = number of requests from other streams that occur between successive requests from this (sequential) stream CAECW-01, 7 4

5 Synthetic experiments to focus on behaviors Synthetic workloads to isolate behaviors: Holes in sequential runs Interleaved streams Evaluation methodology: Compare workload performance with no forced array prefetching vs. forced minimum array prefetching MinPrefetch = 0, 64K (also 16K, 32K, 48K) Performance = array service time no device driver queueing CAECW-01, 8 Workload metrics and real applications Workload metrics Jump distance and interference distance Replay traces from real workloads: 300-GB TPC-D: table, index, temp, log, summary Open Mail server Cello file server traces: root, news, home, ssp Start with single-lun experiments Pick a representative array, and a representative LUN CAECW-01, 9 5

6 Prefetching investigation Exploration of jump distance How does it impact array service time? How big are the jumps in these workloads? Exploration of interference distance How does stream interference impact service time? How much stream interference occurs in workloads? Proposal for new sequentiality workload metric CAECW-01, 10 Impact of jumps on service time? Average Service Tim e (s) R5_4 disk, 4 KB, 50 req/sec, minprefetch=0 1.E+00 1.E+01 1.E+02 1.E+03 1.E+04 1.E+05 1.E+06 1.E+07 Jum p Distance (KB) RunLength=2 x 4K RunLength= 4 x 4K RunLength=16 x 4K RunLength=64 x 4K RunLength=256 x 4K Large runs have consistently low service time Short runs with small jumps (< 64KB) have low service time Disk track buffers? CAECW-01, 11 6

7 Prefetching success for spatial jumps? Average Service Time (s) E+00 1.E+01 1.E+02 1.E+03 1.E+04 1.E+05 1.E+06 1.E+07 Jump Distance (KB) RunLength= 4 x 4K, RunLength=256 x 4K, RunLength=4 x 4K, RunLength=256 x 4K, Prefetching successful for: both long and short runs with small jumps (< ~1MB) CAECW-01, 12 Fraction of accesses How big are workload spatial jumps? E+00 1.E+06 2.E+06 3.E+06 4.E+06 5.E+06 Jump Distance (KB) cello-home cello-news cello-root cello-ssp open-mail tpc-d Cello-ssp and cello-root are mostly sequential Other workloads have short and long jumps forwards (and backwards) CAECW-01, 13 7

8 Impact of interfering streams on service time? Request Rate = 5 req/sec, Average Service Time (s) Number of Streams For small number of streams (<= 20), run length impacts service time Run Length=4 x 4K Run Length=256 x 4K CAECW-01, 14 Prefetching success for interfering streams? Average Service Time (s) Request Rate=5 req/sec, R5_4disk Interference Distance RunLength=4 x 4K, RunLength=256 x 4K, RunLength=4 x 4K, RunLength=256 x 4K, Workload metric to approx. no. of streams: interference distance Prefetching helps for low interference distance (< 50) & large runs Prefetching hurts for all other cases CAECW-01, 15 8

9 Prefetching success for interfering streams (2)? Average Service Time (s) Prefetching helps for low interference distance and large runs true for TPC-D, cello-ssp (sequential accesses) Prefetching hurts for all other cases true for OpenMail, cello-home; not for cello-{root,news} (bursty) CAECW-01, 16 Request Rate=5 req/sec, R5_4disk Interference Distance RunLength=4 x 4K, RunLength=256 x 4K, RunLength=4 x 4K, RunLength=256 x 4K, cello-root, cello-root, cello-home, cello-home, cello-new s, cello-new s, open-mail, tpc-d, cello-ssp, cello-ssp, open-mail, tpc-d, Average Service Time (s) Prefetching success for interfering streams (3)? Interference Distance Focus: constant aggregate request rate Service time increases as # streams increases Prefetching effective for long runs & low aggregate req rates, fewer streams at higher aggregate req rates CAECW-01, 17 R5_4disk ReqRate=500, RunLength=4 x 4K, ReqRate=100, RunLength=4 x 4K, ReqRate=500, RunLength=256 x 4K, ReqRate=100, RunLength=256 x 4K, ReqRate=500, RunLength=4 x 4K, ReqRate=500, RunLength=256 x 4K, ReqRate=100, RunLength=4 x 4K, ReqRate=100, RunLength=256 x 4K, 9

10 Summary of results Array prefetching effective for: Longer runs Small number of streams (low interference distance) For shorter sequential runs, smaller jumps between runs Simple run count spatial locality metrics too restrictive to capture prefetch-friendly behavior New workload characteristics to be captured: Short jumps within a stream Interleaved accesses from other streams Proposal for new run count metric to incorporate these characteristics CAECW-01, 18 Preliminary proposal for sequentiality metric Average Service Time (s) Sequential Run Count (KB) cello-home, cello-home, cello-news, cello-news, cello-root, cello-root, cello-ssp, cello-ssp, open-mail, open-mail, tpc-d, tpc-d, New-and-improved run count: size of run (in KB) where: Short forward jumps permitted Interleaved requests from other streams permitted Does new metric predict performance? Prefetching beneficial for large run counts (TPC-D, cello-ssp) Prefetching ineffective for small run counts (OpenMail) Inconclusive for bursty workloads (cello-{home,news,root}) CAECW-01, 19 10

11 Related work How to present requests to devices, given a workload prefetching and caching (main software differentiator) informed [Tomkins 97], [Kimbrel 96] uninformed [Chervenak 99] Studies of array performance characteristics Few on the real thing: [Chen 90],[ChervenakKatz91] Conseqs of data distribution: layouts [LeeKatz93],[Alvarez 98] How arrays can take advantage of application characteristics Autoraid [Wilkes 96] Setting storage system params:[chenlee95],[chenpatterson90],[shenoyvin97],[jacob96] Commercial products: [EMC s SRDF], [XP s moral eqv] Degraded mode performance studies [HollandGibson93], [ThomasianMenon94] Analytical modeling approaches [Shriver 98],[MuntzLui90] CAECW-01, 20 Conclusions and future work Array features impact performance Developed methodology for evaluating sweet spots Identified several potential sweet spots and behaviors Used synthetic workloads to isolate workload behaviors Measured proposed metrics on traces replayed from real workloads Preliminary results for prefetching behavior for complex workloads New sequentiality metric proposed to capture workload behaviors amenable to prefetching Ongoing/future work Quantify effects of cache capacity and interference Understand consequences of burstiness from synthetic experiments CAECW-01, 21 11

I/O Characterization of Commercial Workloads

I/O Characterization of Commercial Workloads I/O Characterization of Commercial Workloads Kimberly Keeton, Alistair Veitch, Doug Obal, and John Wilkes Storage Systems Program Hewlett-Packard Laboratories www.hpl.hp.com/research/itc/csl/ssp kkeeton@hpl.hp.com

More information

Assert(!Defined(Sequential I/O)) Cheng Li*, Philip Shilane, Fred Douglis, Darren Sawyer, and Hyong Shim

Assert(!Defined(Sequential I/O)) Cheng Li*, Philip Shilane, Fred Douglis, Darren Sawyer, and Hyong Shim Assert(!Defined(Sequential I/O)) Cheng Li*, Philip Shilane, Fred Douglis, Darren Sawyer, and Hyong Shim *utgers University EMC Corporation 1 Sequential I/O is Important Driven by traditional storage characteristics

More information

Getting it Right: Testing Storage Arrays The Way They ll be Used

Getting it Right: Testing Storage Arrays The Way They ll be Used Getting it Right: Testing Storage Arrays The Way They ll be Used Peter Murray Virtual Instruments Flash Memory Summit 2017 Santa Clara, CA 1 The Journey: How Did we Get Here? Storage testing was black

More information

YouChoose: A Performance Interface Enabling Convenient and Efficient QoS Support for Consolidated Storage Systems

YouChoose: A Performance Interface Enabling Convenient and Efficient QoS Support for Consolidated Storage Systems YouChoose: A Performance Interface Enabling Convenient and Efficient QoS Support for Consolidated Storage Systems Xuechen Zhang Yuhai Xu Song Jiang The ECE Department Wayne State University Detroit, MI

More information

Decoupling Datacenter Studies from Access to Large-Scale Applications: A Modeling Approach for Storage Workloads

Decoupling Datacenter Studies from Access to Large-Scale Applications: A Modeling Approach for Storage Workloads Decoupling Datacenter Studies from Access to Large-Scale Applications: A Modeling Approach for Storage Workloads Christina Delimitrou 1, Sriram Sankar 2, Kushagra Vaid 2, Christos Kozyrakis 1 1 Stanford

More information

I/O Commercial Workloads. Scalable Disk Arrays. Scalable ICDA Performance. Outline of This Talk: Related Work on Disk Arrays.

I/O Commercial Workloads. Scalable Disk Arrays. Scalable ICDA Performance. Outline of This Talk: Related Work on Disk Arrays. Scalable Disk Arrays I/O Commercial Workloads David Kaeli Northeastern University Computer Architecture Research Laboratory Boston, MA Manpreet Singh William Zahavi EMC Corporation Hopkington, MA Industry

More information

A Comparison of File. D. Roselli, J. R. Lorch, T. E. Anderson Proc USENIX Annual Technical Conference

A Comparison of File. D. Roselli, J. R. Lorch, T. E. Anderson Proc USENIX Annual Technical Conference A Comparison of File System Workloads D. Roselli, J. R. Lorch, T. E. Anderson Proc. 2000 USENIX Annual Technical Conference File System Performance Integral component of overall system performance Optimised

More information

Database Architecture 2 & Storage. Instructor: Matei Zaharia cs245.stanford.edu

Database Architecture 2 & Storage. Instructor: Matei Zaharia cs245.stanford.edu Database Architecture 2 & Storage Instructor: Matei Zaharia cs245.stanford.edu Summary from Last Time System R mostly matched the architecture of a modern RDBMS» SQL» Many storage & access methods» Cost-based

More information

Erik Riedel Hewlett-Packard Labs

Erik Riedel Hewlett-Packard Labs Erik Riedel Hewlett-Packard Labs Greg Ganger, Christos Faloutsos, Dave Nagle Carnegie Mellon University Outline Motivation Freeblock Scheduling Scheduling Trade-Offs Performance Details Applications Related

More information

DELL TM AX4-5 Application Performance

DELL TM AX4-5 Application Performance DELL TM AX4-5 Application Performance A Comparison of Entry-level Storage Platforms Abstract This paper compares the performance of the Dell AX4-5 with the performance of similarly configured IBM DS3400

More information

PARDA: Proportional Allocation of Resources for Distributed Storage Access

PARDA: Proportional Allocation of Resources for Distributed Storage Access PARDA: Proportional Allocation of Resources for Distributed Storage Access Ajay Gulati, Irfan Ahmad, Carl Waldspurger Resource Management Team VMware Inc. USENIX FAST 09 Conference February 26, 2009 The

More information

A Semi Preemptive Garbage Collector for Solid State Drives. Junghee Lee, Youngjae Kim, Galen M. Shipman, Sarp Oral, Feiyi Wang, and Jongman Kim

A Semi Preemptive Garbage Collector for Solid State Drives. Junghee Lee, Youngjae Kim, Galen M. Shipman, Sarp Oral, Feiyi Wang, and Jongman Kim A Semi Preemptive Garbage Collector for Solid State Drives Junghee Lee, Youngjae Kim, Galen M. Shipman, Sarp Oral, Feiyi Wang, and Jongman Kim Presented by Junghee Lee High Performance Storage Systems

More information

Using EMC FAST with SAP on EMC Unified Storage

Using EMC FAST with SAP on EMC Unified Storage Using EMC FAST with SAP on EMC Unified Storage Applied Technology Abstract This white paper examines the performance considerations of placing SAP applications on FAST-enabled EMC unified storage. It also

More information

A Comprehensive Analytical Performance Model of DRAM Caches

A Comprehensive Analytical Performance Model of DRAM Caches A Comprehensive Analytical Performance Model of DRAM Caches Authors: Nagendra Gulur *, Mahesh Mehendale *, and R Govindarajan + Presented by: Sreepathi Pai * Texas Instruments, + Indian Institute of Science

More information

Generating a Jump Distance Based Synthetic Disk Access Pattern

Generating a Jump Distance Based Synthetic Disk Access Pattern Generating a Jump Distance Based Synthetic Disk Access Pattern Zachary Kurmas Dept. of Computer Science Grand Valley State University kurmasz@gvsu.edu Jeremy Zito, Lucas Trevino, and Ryan Lush Dept. of

More information

Presented by: Nafiseh Mahmoudi Spring 2017

Presented by: Nafiseh Mahmoudi Spring 2017 Presented by: Nafiseh Mahmoudi Spring 2017 Authors: Publication: Type: ACM Transactions on Storage (TOS), 2016 Research Paper 2 High speed data processing demands high storage I/O performance. Flash memory

More information

Performance Modeling and Analysis of Flash based Storage Devices

Performance Modeling and Analysis of Flash based Storage Devices Performance Modeling and Analysis of Flash based Storage Devices H. Howie Huang, Shan Li George Washington University Alex Szalay, Andreas Terzis Johns Hopkins University MSST 11 May 26, 2011 NAND Flash

More information

Warming up Storage-level Caches with Bonfire

Warming up Storage-level Caches with Bonfire Warming up Storage-level Caches with Bonfire Yiying Zhang Gokul Soundararajan Mark W. Storer Lakshmi N. Bairavasundaram Sethuraman Subbiah Andrea C. Arpaci-Dusseau Remzi H. Arpaci-Dusseau 2 Does on-demand

More information

Load Dynamix Enterprise 5.2

Load Dynamix Enterprise 5.2 DATASHEET Load Dynamix Enterprise 5.2 Storage performance analytics for comprehensive workload insight Load DynamiX Enterprise software is the industry s only automated workload acquisition, workload analysis,

More information

Balancing DRAM Locality and Parallelism in Shared Memory CMP Systems

Balancing DRAM Locality and Parallelism in Shared Memory CMP Systems Balancing DRAM Locality and Parallelism in Shared Memory CMP Systems Min Kyu Jeong, Doe Hyun Yoon^, Dam Sunwoo*, Michael Sullivan, Ikhwan Lee, and Mattan Erez The University of Texas at Austin Hewlett-Packard

More information

Oracle on HP Storage

Oracle on HP Storage Oracle on HP Storage Jaime Blasco EMEA HP/Oracle CTC Cooperative Technology Center Boeblingen - November 2004 2003 Hewlett-Packard Development Company, L.P. The information contained herein is subject

More information

Identifying Performance Bottlenecks with Real- World Applications and Flash-Based Storage

Identifying Performance Bottlenecks with Real- World Applications and Flash-Based Storage Identifying Performance Bottlenecks with Real- World Applications and Flash-Based Storage TechTarget Dennis Martin 1 Agenda About Demartek Enterprise Data Center Environments Storage Performance Metrics

More information

Increasing Performance of Existing Oracle RAC up to 10X

Increasing Performance of Existing Oracle RAC up to 10X Increasing Performance of Existing Oracle RAC up to 10X Prasad Pammidimukkala www.gridironsystems.com 1 The Problem Data can be both Big and Fast Processing large datasets creates high bandwidth demand

More information

Technical Note P/N REV A01 March 29, 2007

Technical Note P/N REV A01 March 29, 2007 EMC Symmetrix DMX-3 Best Practices Technical Note P/N 300-004-800 REV A01 March 29, 2007 This technical note contains information on these topics: Executive summary... 2 Introduction... 2 Tiered storage...

More information

IOmark- VM. HP MSA P2000 Test Report: VM a Test Report Date: 4, March

IOmark- VM. HP MSA P2000 Test Report: VM a Test Report Date: 4, March IOmark- VM HP MSA P2000 Test Report: VM- 140304-2a Test Report Date: 4, March 2014 Copyright 2010-2014 Evaluator Group, Inc. All rights reserved. IOmark- VM, IOmark- VDI, VDI- IOmark, and IOmark are trademarks

More information

Proceedings of the FAST 2002 Conference on File and Storage Technologies

Proceedings of the FAST 2002 Conference on File and Storage Technologies USENIX Association Proceedings of the FAST 2002 Conference on File and Storage Technologies Monterey, California, USA January 28-30, 2002 THE ADVANCED COMPUTING SYSTEMS ASSOCIATION 2002 by The USENIX Association

More information

Deduplication File System & Course Review

Deduplication File System & Course Review Deduplication File System & Course Review Kai Li 12/13/13 Topics u Deduplication File System u Review 12/13/13 2 Storage Tiers of A Tradi/onal Data Center $$$$ Mirrored storage $$$ Dedicated Fibre Clients

More information

OASIS: Self-tuning Storage for Applications

OASIS: Self-tuning Storage for Applications OASIS: Self-tuning Storage for Applications Kostas Magoutis, Prasenjit Sarkar, Gauri Shah 14 th NASA Goddard- 23 rd IEEE Mass Storage Systems Technologies, College Park, MD, May 17, 2006 Outline Motivation

More information

EMC CLARiiON Backup Storage Solutions

EMC CLARiiON Backup Storage Solutions Engineering White Paper Backup-to-Disk Guide with Computer Associates BrightStor ARCserve Backup Abstract This white paper describes how to configure EMC CLARiiON CX series storage systems with Computer

More information

Database Workload. from additional misses in this already memory-intensive databases? interference could be a problem) Key question:

Database Workload. from additional misses in this already memory-intensive databases? interference could be a problem) Key question: Database Workload + Low throughput (0.8 IPC on an 8-wide superscalar. 1/4 of SPEC) + Naturally threaded (and widely used) application - Already high cache miss rates on a single-threaded machine (destructive

More information

Performance measurement. SMD149 - Operating Systems - Performance and processor design. Introduction. Important trends affecting performance issues

Performance measurement. SMD149 - Operating Systems - Performance and processor design. Introduction. Important trends affecting performance issues Performance measurement SMD149 - Operating Systems - Performance and processor design Roland Parviainen November 28, 2005 Performance measurement Motivation Techniques Common metrics Processor architectural

More information

Evaluating the Impact of RDMA on Storage I/O over InfiniBand

Evaluating the Impact of RDMA on Storage I/O over InfiniBand Evaluating the Impact of RDMA on Storage I/O over InfiniBand J Liu, DK Panda and M Banikazemi Computer and Information Science IBM T J Watson Research Center The Ohio State University Presentation Outline

More information

Chapter 14 Performance and Processor Design

Chapter 14 Performance and Processor Design Chapter 14 Performance and Processor Design Outline 14.1 Introduction 14.2 Important Trends Affecting Performance Issues 14.3 Why Performance Monitoring and Evaluation are Needed 14.4 Performance Measures

More information

Today s Papers. Array Reliability. RAID Basics (Two optional papers) EECS 262a Advanced Topics in Computer Systems Lecture 3

Today s Papers. Array Reliability. RAID Basics (Two optional papers) EECS 262a Advanced Topics in Computer Systems Lecture 3 EECS 262a Advanced Topics in Computer Systems Lecture 3 Filesystems (Con t) September 10 th, 2012 John Kubiatowicz and Anthony D. Joseph Electrical Engineering and Computer Sciences University of California,

More information

My cache or yours? Making storage more exclusive

My cache or yours? Making storage more exclusive USENIXAnnual Technical Conference (USENIX2002), pp. 161 175, 10 15 June 2002, Monterey,CA.(USENIX,Berkeley,CA.) My cache or yours? Making storage more exclusive Theodore M. Wong Carnegie Mellon University,

More information

An Introduction to I/O and Storage Tuning. Randy Kreiser Senior MTS Consulting Engineer, SGI

An Introduction to I/O and Storage Tuning. Randy Kreiser Senior MTS Consulting Engineer, SGI Randy Kreiser Senior MTS Consulting Engineer, SGI 40/30/30 Performance Rule 40% Hardware Setup 30% System Software Setup 30% Application Software Analyze the Application Large/Small I/O s Sequential/Random

More information

BITUG Big Sig December 2013 NonStop Performance Update David Sly HP UK Tech Services

BITUG Big Sig December 2013 NonStop Performance Update David Sly HP UK Tech Services BITUG Big Sig December 2013 NonStop Performance Update David Sly HP UK Tech Services 1 HP confidential information This is a rolling (up to three year) Roadmap and is subject to change without notice.

More information

GENIUS: Generator of Interactive User Media Sessions

GENIUS: Generator of Interactive User Media Sessions GENIUS: Generator of Interactive User Media Sessions Claudiney Ramos Cristiano Costa Ítalo Cunha Jussara M. Almeida Department of Computer Science Federal University of Minas Gerais Brazil Motivation Realistic

More information

Analyzing Memory Access Patterns and Optimizing Through Spatial Memory Streaming. Ogün HEPER CmpE 511 Computer Architecture December 24th, 2009

Analyzing Memory Access Patterns and Optimizing Through Spatial Memory Streaming. Ogün HEPER CmpE 511 Computer Architecture December 24th, 2009 Analyzing Memory Access Patterns and Optimizing Through Spatial Memory Streaming Ogün HEPER CmpE 511 Computer Architecture December 24th, 2009 Agenda Introduction Memory Hierarchy Design CPU Speed vs.

More information

The Oracle Database Appliance I/O and Performance Architecture

The Oracle Database Appliance I/O and Performance Architecture Simple Reliable Affordable The Oracle Database Appliance I/O and Performance Architecture Tammy Bednar, Sr. Principal Product Manager, ODA 1 Copyright 2012, Oracle and/or its affiliates. All rights reserved.

More information

Tradeoff between coverage of a Markov prefetcher and memory bandwidth usage

Tradeoff between coverage of a Markov prefetcher and memory bandwidth usage Tradeoff between coverage of a Markov prefetcher and memory bandwidth usage Elec525 Spring 2005 Raj Bandyopadhyay, Mandy Liu, Nico Peña Hypothesis Some modern processors use a prefetching unit at the front-end

More information

Performance Testing December 16, 2017

Performance Testing December 16, 2017 December 16, 2017 1 1. vsan Performance Testing 1.1.Performance Testing Overview Table of Contents 2 1. vsan Performance Testing Performance Testing 3 1.1 Performance Testing Overview Performance Testing

More information

W H I T E P A P E R. Comparison of Storage Protocol Performance in VMware vsphere 4

W H I T E P A P E R. Comparison of Storage Protocol Performance in VMware vsphere 4 W H I T E P A P E R Comparison of Storage Protocol Performance in VMware vsphere 4 Table of Contents Introduction................................................................... 3 Executive Summary............................................................

More information

Buttress: A toolkit for flexible and high fidelity I/O benchmarking

Buttress: A toolkit for flexible and high fidelity I/O benchmarking Appears in the Proceedings of the 3rd Conference on File and Storage Technologies (FAST 4). (31 Mar 2 Apr 24, San Francisco, CA). Published by USENIX, Berkeley, CA. Buttress: A toolkit for flexible and

More information

Forget IOPS: A Proper Way to Characterize & Test Storage Performance Peter Murray SwiftTest

Forget IOPS: A Proper Way to Characterize & Test Storage Performance Peter Murray SwiftTest Forget IOPS: A Proper Way to Characterize & Test Storage Performance Peter Murray peter@swifttest.com SwiftTest Storage Performance Validation Rely on vendor IOPS claims Test in production and pray Validate

More information

Issues and Challenges in the Performance Analysis of Real Disk Arrays

Issues and Challenges in the Performance Analysis of Real Disk Arrays IEEE TRANSACTIONS ON PARALLEL AND DISTRIBUTED SYSTEMS, VOL. 15, NO. 6, JUNE 2004 559 Issues and Challenges in the Performance Analysis of Real Disk Arrays Elizabeth Varki, Member, IEEE, Arif Merchant,

More information

Using Synology SSD Technology to Enhance System Performance Synology Inc.

Using Synology SSD Technology to Enhance System Performance Synology Inc. Using Synology SSD Technology to Enhance System Performance Synology Inc. Synology_WP_ 20121112 Table of Contents Chapter 1: Enterprise Challenges and SSD Cache as Solution Enterprise Challenges... 3 SSD

More information

Feng Chen and Xiaodong Zhang Dept. of Computer Science and Engineering The Ohio State University

Feng Chen and Xiaodong Zhang Dept. of Computer Science and Engineering The Ohio State University Caching for Bursts (C-Burst): Let Hard Disks Sleep Well and Work Energetically Feng Chen and Xiaodong Zhang Dept. of Computer Science and Engineering The Ohio State University Power Management in Hard

More information

Typical scenario in shared infrastructures

Typical scenario in shared infrastructures Got control? AutoControl: Automated Control of MultipleVirtualized Resources Pradeep Padala, Karen Hou, Xiaoyun Zhu*, Mustfa Uysal, Zhikui Wang, Sharad Singhal, Arif Merchant, Kang G. Shin University of

More information

NetApp AFF A300 Review

NetApp AFF A300 Review StorageReview StorageReview takes an in-depth look at features, and performance of NetApp AFF A300 storage array. 1395 Crossman Ave Sunnyvale, CA 94089 United States Table of Contents INTRODUCTION... 3-5

More information

IOmark- VDI. IBM IBM FlashSystem V9000 Test Report: VDI a Test Report Date: 5, December

IOmark- VDI. IBM IBM FlashSystem V9000 Test Report: VDI a Test Report Date: 5, December IOmark- VDI IBM IBM FlashSystem V9000 Test Report: VDI- 151205- a Test Report Date: 5, December 2015 Copyright 2010-2015 Evaluator Group, Inc. All rights reserved. IOmark- VDI, IOmark- VM, VDI- IOmark,

More information

Dell EMC Unity: Performance Analysis Deep Dive. Keith Snell Performance Engineering Midrange & Entry Solutions Group

Dell EMC Unity: Performance Analysis Deep Dive. Keith Snell Performance Engineering Midrange & Entry Solutions Group Dell EMC Unity: Performance Analysis Deep Dive Keith Snell Performance Engineering Midrange & Entry Solutions Group Agenda Introduction Sample Period Unisphere Performance Dashboard Unisphere uemcli command

More information

VERITAS Storage Foundation 4.0 for Oracle

VERITAS Storage Foundation 4.0 for Oracle D E C E M B E R 2 0 0 4 VERITAS Storage Foundation 4.0 for Oracle Performance Brief AIX 5.2, Oracle 9iR2 VERITAS Storage Foundation for Oracle Abstract This document details the high performance characteristics

More information

Managing Performance Variance of Applications Using Storage I/O Control

Managing Performance Variance of Applications Using Storage I/O Control Performance Study Managing Performance Variance of Applications Using Storage I/O Control VMware vsphere 4.1 Application performance can be impacted when servers contend for I/O resources in a shared storage

More information

EMC Tiered Storage for Microsoft SQL Server 2008 Enabled by EMC CLARiiON CX4 and Enterprise Flash Drives

EMC Tiered Storage for Microsoft SQL Server 2008 Enabled by EMC CLARiiON CX4 and Enterprise Flash Drives EMC Tiered Storage for Microsoft SQL Server 2008 Enabled by EMC CLARiiON CX4 and Enterprise Flash Drives A Detailed Review EMC Information Infrastructure Solutions Abstract This white paper demonstrates

More information

Adaptive disk scheduling for overload management

Adaptive disk scheduling for overload management Adaptive disk scheduling for overload management Alma Riska Seagate Research 1251 Waterfront Place Pittsburgh, PA 15222 Alma.Riska@seagate.com Erik Riedel Seagate Research 1251 Waterfront Place Pittsburgh,

More information

ASEP: An Adaptive Sequential Prefetching Scheme for Second-level Storage System

ASEP: An Adaptive Sequential Prefetching Scheme for Second-level Storage System ASEP: An Adaptive Sequential Prefetching Scheme for Second-level Storage System Xiaodong Shi Email: shixd.hust@gmail.com Dan Feng Email: dfeng@hust.edu.cn Wuhan National Laboratory for Optoelectronics,

More information

Design Tradeoffs for Data Deduplication Performance in Backup Workloads

Design Tradeoffs for Data Deduplication Performance in Backup Workloads Design Tradeoffs for Data Deduplication Performance in Backup Workloads Min Fu,DanFeng,YuHua,XubinHe, Zuoning Chen *, Wen Xia,YuchengZhang,YujuanTan Huazhong University of Science and Technology Virginia

More information

Dell PowerEdge R720xd with PERC H710P: A Balanced Configuration for Microsoft Exchange 2010 Solutions

Dell PowerEdge R720xd with PERC H710P: A Balanced Configuration for Microsoft Exchange 2010 Solutions Dell PowerEdge R720xd with PERC H710P: A Balanced Configuration for Microsoft Exchange 2010 Solutions A comparative analysis with PowerEdge R510 and PERC H700 Global Solutions Engineering Dell Product

More information

Isilon Performance. Name

Isilon Performance. Name 1 Isilon Performance Name 2 Agenda Architecture Overview Next Generation Hardware Performance Caching Performance Streaming Reads Performance Tuning OneFS Architecture Overview Copyright 2014 EMC Corporation.

More information

Statistical Simulation of Superscalar Architectures using Commercial Workloads

Statistical Simulation of Superscalar Architectures using Commercial Workloads Statistical Simulation of Superscalar Architectures using Commercial Workloads Lieven Eeckhout and Koen De Bosschere Dept. of Electronics and Information Systems (ELIS) Ghent University, Belgium CAECW

More information

Improving DRAM Performance by Parallelizing Refreshes with Accesses

Improving DRAM Performance by Parallelizing Refreshes with Accesses Improving DRAM Performance by Parallelizing Refreshes with Accesses Kevin Chang Donghyuk Lee, Zeshan Chishti, Alaa Alameldeen, Chris Wilkerson, Yoongu Kim, Onur Mutlu Executive Summary DRAM refresh interferes

More information

AN ALTERNATIVE TO ALL- FLASH ARRAYS: PREDICTIVE STORAGE CACHING

AN ALTERNATIVE TO ALL- FLASH ARRAYS: PREDICTIVE STORAGE CACHING AN ALTERNATIVE TO ALL- FLASH ARRAYS: PREDICTIVE STORAGE CACHING THE EASIEST WAY TO INCREASE PERFORMANCE AND LOWER STORAGE COSTS Bruce Kornfeld, Chief Marketing Officer, StorMagic Luke Pruen, Technical

More information

Assessing performance in HP LeftHand SANs

Assessing performance in HP LeftHand SANs Assessing performance in HP LeftHand SANs HP LeftHand Starter, Virtualization, and Multi-Site SANs deliver reliable, scalable, and predictable performance White paper Introduction... 2 The advantages of

More information

EMC Disk Tiering Technology Review

EMC Disk Tiering Technology Review EMC Disk Tiering Technology Review Tony Negro EMC Corporation Wednesday, February 06, 2013 12:15 PM Session Number 13154 Agenda Basis for FAST Implementation Characteristics Operational Considerations

More information

Design Considerations for Using Flash Memory for Caching

Design Considerations for Using Flash Memory for Caching Design Considerations for Using Flash Memory for Caching Edi Shmueli, IBM XIV Storage Systems edi@il.ibm.com Santa Clara, CA August 2010 1 Solid-State Storage In a few decades solid-state storage will

More information

Data Storage Institute. SANSIM: A PLATFORM FOR SIMULATION AND DESIGN OF A STORAGE AREA NETWORK Zhu Yaolong

Data Storage Institute. SANSIM: A PLATFORM FOR SIMULATION AND DESIGN OF A STORAGE AREA NETWORK Zhu Yaolong Data Storage Institute SANSIM: A PLATFORM FOR SIMULATION AND DESIGN OF A STORAGE AREA NETWORK Zhu Yaolong e_mail:zhu_yaolong@dsi.a-star.edu.sg Outline Motivation Key Focuses Simulation Methodology SANSim

More information

Bridging the Processor/Memory Performance Gap in Database Applications

Bridging the Processor/Memory Performance Gap in Database Applications Bridging the Processor/Memory Performance Gap in Database Applications Anastassia Ailamaki Carnegie Mellon http://www.cs.cmu.edu/~natassa Memory Hierarchies PROCESSOR EXECUTION PIPELINE L1 I-CACHE L1 D-CACHE

More information

Generating a Jump Distance Based Synthetic Disk Access Pattern

Generating a Jump Distance Based Synthetic Disk Access Pattern Grand Valley State University ScholarWorks@GVSU Technical Library School of Computing and Information Systems 2006 Generating a Jump Distance Based Synthetic Disk Access Pattern Dr. Zachary Kurmas Grand

More information

Hippodrome: running circles around storage administration

Hippodrome: running circles around storage administration Conference on File and Storage Technologies (FAST 2), pp. 175 188, 28 3 January 22, Monterey, CA. (USENIX, Berkeley, CA.) Hippodrome: running circles around storage administration Eric Anderson, Michael

More information

Mesocode: Optimizations for Improving Fetch Bandwidth of Future Itanium Processors

Mesocode: Optimizations for Improving Fetch Bandwidth of Future Itanium Processors : Optimizations for Improving Fetch Bandwidth of Future Itanium Processors Marsha Eng, Hong Wang, Perry Wang Alex Ramirez, Jim Fung, and John Shen Overview Applications of for Itanium Improving fetch bandwidth

More information

CSE544 Database Architecture

CSE544 Database Architecture CSE544 Database Architecture Tuesday, February 1 st, 2011 Slides courtesy of Magda Balazinska 1 Where We Are What we have already seen Overview of the relational model Motivation and where model came from

More information

Optimizing Flash-based Key-value Cache Systems

Optimizing Flash-based Key-value Cache Systems Optimizing Flash-based Key-value Cache Systems Zhaoyan Shen, Feng Chen, Yichen Jia, Zili Shao Department of Computing, Hong Kong Polytechnic University Computer Science & Engineering, Louisiana State University

More information

Data Storage and Query Answering. Data Storage and Disk Structure (2)

Data Storage and Query Answering. Data Storage and Disk Structure (2) Data Storage and Query Answering Data Storage and Disk Structure (2) Review: The Memory Hierarchy Swapping, Main-memory DBMS s Tertiary Storage: Tape, Network Backup 3,200 MB/s (DDR-SDRAM @200MHz) 6,400

More information

MQSim: A Framework for Enabling Realistic Studies of Modern Multi-Queue SSD Devices

MQSim: A Framework for Enabling Realistic Studies of Modern Multi-Queue SSD Devices MQSim: A Framework for Enabling Realistic Studies of Modern Multi-Queue SSD Devices Arash Tavakkol, Juan Gómez-Luna, Mohammad Sadrosadati, Saugata Ghose, Onur Mutlu February 13, 2018 Executive Summary

More information

ibench: Quantifying Interference in Datacenter Applications

ibench: Quantifying Interference in Datacenter Applications ibench: Quantifying Interference in Datacenter Applications Christina Delimitrou and Christos Kozyrakis Stanford University IISWC September 23 th 2013 Executive Summary Problem: Increasing utilization

More information

A Case Study: Performance Evaluation of a DRAM-Based Solid State Disk

A Case Study: Performance Evaluation of a DRAM-Based Solid State Disk A Case Study: Performance Evaluation of a DRAM-Based Solid State Disk Hitoshi Oi The University of Aizu November 2, 2007 Japan-China Joint Workshop on Frontier of Computer Science and Technology (FCST)

More information

Operating Systems. Lecture File system implementation. Master of Computer Science PUF - Hồ Chí Minh 2016/2017

Operating Systems. Lecture File system implementation. Master of Computer Science PUF - Hồ Chí Minh 2016/2017 Operating Systems Lecture 7.2 - File system implementation Adrien Krähenbühl Master of Computer Science PUF - Hồ Chí Minh 2016/2017 Design FAT or indexed allocation? UFS, FFS & Ext2 Journaling with Ext3

More information

Toward Energy-efficient and Fault-tolerant Consistent Hashing based Data Store. Wei Xie TTU CS Department Seminar, 3/7/2017

Toward Energy-efficient and Fault-tolerant Consistent Hashing based Data Store. Wei Xie TTU CS Department Seminar, 3/7/2017 Toward Energy-efficient and Fault-tolerant Consistent Hashing based Data Store Wei Xie TTU CS Department Seminar, 3/7/2017 1 Outline General introduction Study 1: Elastic Consistent Hashing based Store

More information

SoftNAS Cloud Performance Evaluation on Microsoft Azure

SoftNAS Cloud Performance Evaluation on Microsoft Azure SoftNAS Cloud Performance Evaluation on Microsoft Azure November 30, 2016 Contents SoftNAS Cloud Overview... 3 Introduction... 3 Executive Summary... 4 Key Findings for Azure:... 5 Test Methodology...

More information

Recommendations for Aligning VMFS Partitions

Recommendations for Aligning VMFS Partitions VMWARE PERFORMANCE STUDY VMware ESX Server 3.0 Recommendations for Aligning VMFS Partitions Partition alignment is a known issue in physical file systems, and its remedy is well-documented. The goal of

More information

Recovering Disk Storage Metrics from low level Trace events

Recovering Disk Storage Metrics from low level Trace events Recovering Disk Storage Metrics from low level Trace events Progress Report Meeting May 05, 2016 Houssem Daoud Michel Dagenais École Polytechnique de Montréal Laboratoire DORSAL Agenda Introduction and

More information

Key Point. What are Cache lines

Key Point. What are Cache lines Caching 1 Key Point What are Cache lines Tags Index offset How do we find data in the cache? How do we tell if it s the right data? What decisions do we need to make in designing a cache? What are possible

More information

Two hours - online. The exam will be taken on line. This paper version is made available as a backup

Two hours - online. The exam will be taken on line. This paper version is made available as a backup COMP 25212 Two hours - online The exam will be taken on line. This paper version is made available as a backup UNIVERSITY OF MANCHESTER SCHOOL OF COMPUTER SCIENCE System Architecture Date: Monday 21st

More information

Thread Cluster Memory Scheduling: Exploiting Differences in Memory Access Behavior. Yoongu Kim Michael Papamichael Onur Mutlu Mor Harchol-Balter

Thread Cluster Memory Scheduling: Exploiting Differences in Memory Access Behavior. Yoongu Kim Michael Papamichael Onur Mutlu Mor Harchol-Balter Thread Cluster Memory Scheduling: Exploiting Differences in Memory Access Behavior Yoongu Kim Michael Papamichael Onur Mutlu Mor Harchol-Balter Motivation Memory is a shared resource Core Core Core Core

More information

CS 147: Computer Systems Performance Analysis

CS 147: Computer Systems Performance Analysis CS 147: Computer Systems Performance Analysis Test Loads CS 147: Computer Systems Performance Analysis Test Loads 1 / 33 Overview Overview Overview 2 / 33 Test Load Design Test Load Design Test Load Design

More information

Evaluation Report: HP StoreFabric SN1000E 16Gb Fibre Channel HBA

Evaluation Report: HP StoreFabric SN1000E 16Gb Fibre Channel HBA Evaluation Report: HP StoreFabric SN1000E 16Gb Fibre Channel HBA Evaluation report prepared under contract with HP Executive Summary The computing industry is experiencing an increasing demand for storage

More information

HP AutoRAID (Lecture 5, cs262a)

HP AutoRAID (Lecture 5, cs262a) HP AutoRAID (Lecture 5, cs262a) Ion Stoica, UC Berkeley September 13, 2016 (based on presentation from John Kubiatowicz, UC Berkeley) Array Reliability Reliability of N disks = Reliability of 1 Disk N

More information

Virtualizing SQL Server 2008 Using EMC VNX Series and VMware vsphere 4.1. Reference Architecture

Virtualizing SQL Server 2008 Using EMC VNX Series and VMware vsphere 4.1. Reference Architecture Virtualizing SQL Server 2008 Using EMC VNX Series and VMware vsphere 4.1 Copyright 2011, 2012 EMC Corporation. All rights reserved. Published March, 2012 EMC believes the information in this publication

More information

Copyright 2012, Elsevier Inc. All rights reserved.

Copyright 2012, Elsevier Inc. All rights reserved. Computer Architecture A Quantitative Approach, Fifth Edition Chapter 2 Memory Hierarchy Design 1 Introduction Introduction Programmers want unlimited amounts of memory with low latency Fast memory technology

More information

Threshold-Based Markov Prefetchers

Threshold-Based Markov Prefetchers Threshold-Based Markov Prefetchers Carlos Marchani Tamer Mohamed Lerzan Celikkanat George AbiNader Rice University, Department of Electrical and Computer Engineering ELEC 525, Spring 26 Abstract In this

More information

In-Memory Data Management

In-Memory Data Management In-Memory Data Management Martin Faust Research Assistant Research Group of Prof. Hasso Plattner Hasso Plattner Institute for Software Engineering University of Potsdam Agenda 2 1. Changed Hardware 2.

More information

Computer Architecture. A Quantitative Approach, Fifth Edition. Chapter 2. Memory Hierarchy Design. Copyright 2012, Elsevier Inc. All rights reserved.

Computer Architecture. A Quantitative Approach, Fifth Edition. Chapter 2. Memory Hierarchy Design. Copyright 2012, Elsevier Inc. All rights reserved. Computer Architecture A Quantitative Approach, Fifth Edition Chapter 2 Memory Hierarchy Design 1 Programmers want unlimited amounts of memory with low latency Fast memory technology is more expensive per

More information

EITF20: Computer Architecture Part4.1.1: Cache - 2

EITF20: Computer Architecture Part4.1.1: Cache - 2 EITF20: Computer Architecture Part4.1.1: Cache - 2 Liang Liu liang.liu@eit.lth.se 1 Outline Reiteration Cache performance optimization Bandwidth increase Reduce hit time Reduce miss penalty Reduce miss

More information

Memory Hierarchy Computing Systems & Performance MSc Informatics Eng. Memory Hierarchy (most slides are borrowed)

Memory Hierarchy Computing Systems & Performance MSc Informatics Eng. Memory Hierarchy (most slides are borrowed) Computing Systems & Performance Memory Hierarchy MSc Informatics Eng. 2011/12 A.J.Proença Memory Hierarchy (most slides are borrowed) AJProença, Computer Systems & Performance, MEI, UMinho, 2011/12 1 2

More information

Database Services at CERN with Oracle 10g RAC and ASM on Commodity HW

Database Services at CERN with Oracle 10g RAC and ASM on Commodity HW Database Services at CERN with Oracle 10g RAC and ASM on Commodity HW UKOUG RAC SIG Meeting London, October 24 th, 2006 Luca Canali, CERN IT CH-1211 LCGenève 23 Outline Oracle at CERN Architecture of CERN

More information

Making Storage Smarter Jim Williams Martin K. Petersen

Making Storage Smarter Jim Williams Martin K. Petersen Making Storage Smarter Jim Williams Martin K. Petersen Agenda r Background r Examples r Current Work r Future 2 Definition r Storage is made smarter by exchanging information between the application and

More information

HP AutoRAID (Lecture 5, cs262a)

HP AutoRAID (Lecture 5, cs262a) HP AutoRAID (Lecture 5, cs262a) Ali Ghodsi and Ion Stoica, UC Berkeley January 31, 2018 (based on slide from John Kubiatowicz, UC Berkeley) Array Reliability Reliability of N disks = Reliability of 1 Disk

More information

Performance Benefits of NVMe over Fibre Channel A New, Parallel, Efficient Protocol

Performance Benefits of NVMe over Fibre Channel A New, Parallel, Efficient Protocol May 2018 Performance Benefits of NVMe over Fibre Channel A New, Parallel, Efficient NVMe over Fibre Channel delivered 58% higher IOPS and 34% lower latency than SCSI FCP. (What s not to like?) Executive

More information

Memory Hierarchy Computing Systems & Performance MSc Informatics Eng. Memory Hierarchy (most slides are borrowed)

Memory Hierarchy Computing Systems & Performance MSc Informatics Eng. Memory Hierarchy (most slides are borrowed) Computing Systems & Performance Memory Hierarchy MSc Informatics Eng. 2012/13 A.J.Proença Memory Hierarchy (most slides are borrowed) AJProença, Computer Systems & Performance, MEI, UMinho, 2012/13 1 2

More information