DBMS Data Loading: An Analysis on Modern Hardware. Adam Dziedzic, Manos Karpathiotakis*, Ioannis Alagiannis, Raja Appuswamy, Anastasia Ailamaki
|
|
|
- Matthew Higgins
- 6 years ago
- Views:
Transcription
1 DBMS Data Loading: An Analysis on Modern Hardware Adam Dziedzic, Manos Karpathiotakis*, Ioannis Alagiannis, Raja Appuswamy, Anastasia Ailamaki
2 Data loading: A necessary evil Volume => Expensive 4 zettabytes by 22* Velocity => Continuous Fresh data = Interesting data Top query performance ACID guarantees * [IDC12] Abad [IISWC12] Data loading is a persistent analysis bottleneck 2
3 Loading a DBMS Convert Read Load How does hardware affect loading? 3
4 Loading a DBMS Convert Read Load How does hardware affect loading? 3
5 Hardware Experimental setup Dual socket 8 cores Intel(R) Xeon(R) CPU E GB RAM HDD: 4 x 5 GB 7.5k RPM SATA disks SSD: 3 x 2GB SSD disks DAS: 24 x 5 GB 7.5k RPM SATA disks Software PostgreSQL, DBMS-R MonetDB, DBMS-C PostgreSQL parallel external loader ( PCOPY ) Benchmarks & Real-world Datasets 4
6 Loading Time (sec) Loading Time (sec) Single-threaded data loading [Input storage: HDD Destination storage: DAS] TPC-H Loading Time Symantec Loading Time 75 6 DBMS-C PostgreSQL MonetDB DBMS-R File size (GB) File size (GB) Dataset characteristics matter Effect of compression 5
7 Loading Time (sec) Loading Time (sec) Parallel data loading Input storage: HDD - Destination storage: DAS 16 threads TPC-H Loading Time Symantec Loading Time 3 DBMS-C 3 2 MonetDB DBMS-R 2 1 PCOPY File size (GB) File size (GB) Speedup 16 threads DBMS-R PCOPY MonetDB DBMS-C TPC-H 1GB Symantec 1GB Sublinear speedup for 16 threads 6
8 Utilzation (%) Resource Utilization DBMS-R, TPC-H SF1 In: HDD - Out: DAS Time (sec) Unable to saturate resources Write BW 7
9 Utilzation (%) Resource Utilization DBMS-R, TPC-H SF1 In: HDD - Out: DAS Time (sec) Unable to saturate resources Write BW 7
10 Utilzation (%) Resource Utilization DBMS-R, TPC-H SF1 In: HDD - Out: DAS Time (sec) Read BW Unable to saturate resources Write BW 7
11 Utilzation (%) Resource Utilization DBMS-R, TPC-H SF1 In: HDD - Out: DAS Time (sec) Unable to saturate resources CPU Read BW Write BW 7
12 Utilzation (%) I/O Wait (%) Resource Utilization DBMS-R MonetDB DBMS-R, TPC-H SF1 In: HDD - Out: DAS Time (sec) Unable to saturate resources CPU Read BW Write BW 7
13 Block Address Block Address Read patterns [TPC-H SF1 Input storage: HDD Destination storage: DAS] MonetDB DBMS-R Elapsed Time Elapsed Time Random I/O causes underutilization 8
14 Read Utilzation (%) Serial reader vs. Parallel readers [TPC-H SF1 Input storage: HDD Destination storage: DAS] PCOPY read utilization Parallel Readers Time (sec) Serial Reader Serial reader improves read utilization # readers depends on input device speed 9
15 Impact of storage 1
16 Loading Time (sec) Loading Time (sec) Impact of storage [TPC-H SF1] Slow input storage Varying input storage HDD Source Storage DAS Destination Storage 4 HDD to DAS 4 HDD to DAS 3 2 HDD to SSD HDD to ramfs 3 2 SSD to DAS ramfs to DAS 1 1 PCOPY DBMS-R DBMS-C PCOPY DBMS-R DBMS-C Slow source storage bottlenecks all systems Write bottleneck when source storage is fast 11
17 Loading Time (sec) CPU Utilization (%) Best-case storage scenario [TPC-H SF1] ramfs Source Storage ramfs to HDD ramfs to DAS DBMS-R In: ramfs Out: ramfs 4 ramfs to SSD ramfs to ramfs PCOPY DBMS-R DBMS-C Time (sec) Device Bandwidth: 12.8 GB/sec Read Rate: 25 MB/sec 1% CPU utilization, yet B/W still underutilized 12
18 CPU breakdown (%) Data loading: Where does time go? 1% 8% 6% 4% [1 int columns; 1GB] Other Write data Logging Tuple creation Conversion 2% % PostgreSQL MonetDB Tokenizing Parsing Parsing, conversion, tokenization hotspots 13
19 Reducing data loading overheads SDS/Q Reduced Accesses NoDB Data Vaults RAW Optimized Code Path Instant Loading HW FPGAs GPUs 14
20 Bulk loading on modern hardware General case: Resource under-utilization Slow destination storage matters Complex code paths bound max speed 15
21 Bulk loading on modern hardware General case: Resource under-utilization Slow destination storage matters Complex code paths bound max speed Thank You! Questions? 15
22 Backup Slides 16
![5x data growth from 21 to 22 [IDC212]](/docs-images/72/67906736/images/23-0.jpg "Can DBMS keep up with data growth? 23")
23 5x data growth from 21 to 22 [IDC212] Can DBMS keep up with data growth? 23
24 Storage Characteristics Name Capacity Configuration Read Speed Write Speed RPM HDD 2TB 4 x HDD (RAID-) 17 MB/s 16 MB/s 7.5K DAS 12TB 24 x HDD (RAID-) 11 MB/s 33 MB/s 7.5K SSD 6GB 3 x SSD (RAID-) 565 MB/s 268 MB/s n/a 24
25 Speedup Speedup Parallel data loading 16 threads [Input storage: HDD Destination storage: DAS] TPC-H Symantec PCOPY DBMS-R MonetDB DBMS-C File size (GB) File size (GB) Sublinear speedup for 16x DoP 7
26 Loading Time (sec) Loading Time (sec) Single-threaded loading Extra datasets TPC-C Loading Time PostgreSQL MonetDB DBMS-R DBMS-C Input storage: HDD Destination storage: DAS SDSS Loading Time File size (GB) File size (GB) Column stores invest in compression 26
27 Loading Time (sec) Loading Time (sec) Parallel data loading Extra datasets Input storage: HDD Destination storage: DAS TPC-C Loading Time SDSS Loading Time 4 PCOPY MonetDB DBMS-R DBMS-C File size (GB) File size (GB) 27
28 Speedup Speedup Parallel data loading Extra datasets Input storage: HDD Destination storage: DAS TPC-C SDSS 4 3 PostgreSQL DBMS-R MonetDB DBMS-C File size (GB) File size (GB) 28
29 The effect of compression [1GB] DB size / input file Name TPC-H TPC-C SDSS Symantec DBMS-R PostgreSQL DBMS-C MonetDB Column stores: Reduced footprint favors OLAP 29
30 Utilzation (%) I/O Wait (%) Resource Utilization PCOPY DBMS-R MonetDB DBMS-C DBMS-R, TPC-H SF1 In: HDD - Out: DAS Time (sec) Unable to saturate resources CPU Read BW Write BW 8
31 Utilzation (%) MonetDB utilization [Data: TPCH SF1 Input storage: HDD Destination storage: DAS] Time (sec) 31
32 Utilzation (%) PCOPY utilization [Data: TPCH SF1 Input storage: HDD Destination storage: DAS] CPU Read BW Time (sec) 32
33 Utilzation (%) DBMS-C utilization [Data: TPCH SF1 Input storage: HDD Destination storage: DAS] Time (sec) 33
34 DBMS-C read patterns 34
35 Reducing data loading overheads In situ querying [SIGMOD12, VLDB14] Data Vaults: Exploit metadata [Ivanova12,Kargin15] Instant Loading: SIMD & Code gen. [Muehlbauer13] Accelerators (FPGAs, GPUs) 35
DBMS Data Loading: An Analysis on Modern Hardware
DBMS Data Loading: An Analysis on Modern Hardware Adam Dziedzic 1, Manos Karpathiotakis 2, Ioannis Alagiannis 2, Raja Appuswamy 2, and Anastasia Ailamaki 2,3 1 University of Chicago ady@uchicago.edu 2
Data Transformation and Migration in Polystores
Data Transformation and Migration in Polystores Adam Dziedzic, Aaron Elmore & Michael Stonebraker September 15th, 2016 Agenda Data Migration for Polystores: What & Why? How? Acceleration of physical data
The Case for Heterogeneous HTAP
The Case for Heterogeneous HTAP Raja Appuswamy, Manos Karpathiotakis, Danica Porobic, and Anastasia Ailamaki Data-Intensive Applications and Systems Lab EPFL 1 HTAP the contract with the hardware Hybrid
NoDB: Querying Raw Data. --Mrutyunjay
NoDB: Querying Raw Data --Mrutyunjay Overview Introduction Motivation NoDB Philosophy: PostgreSQL Results Opportunities NoDB in Action: Adaptive Query Processing on Raw Data Ioannis Alagiannis, Renata
Parallel In-situ Data Processing Techniques
Parallel In-situ Data Processing Techniques Florin Rusu, Yu Cheng University of California, Merced Outline Background SCANRAW Operator Speculative Loading Evaluation Astronomy: FITS Format Sloan Digital
Toward timely, predictable and cost-effective data analytics. Renata Borovica-Gajić DIAS, EPFL
Toward timely, predictable and cost-effective data analytics Renata Borovica-Gajić DIAS, EPFL Big data proliferation Big data is when the current technology does not enable users to obtain timely, cost-effective,
Join Processing for Flash SSDs: Remembering Past Lessons
Join Processing for Flash SSDs: Remembering Past Lessons Jaeyoung Do, Jignesh M. Patel Department of Computer Sciences University of Wisconsin-Madison $/MB GB Flash Solid State Drives (SSDs) Benefits of
Parallel In-situ Data Processing with Speculative Loading
Parallel In-situ Data Processing with Speculative Loading Yu Cheng, Florin Rusu University of California, Merced Outline Background Scanraw Operator Speculative Loading Evaluation SAM/BAM Format More than
Smooth Scan: Statistics-Oblivious Access Paths. Renata Borovica-Gajic Stratos Idreos Anastasia Ailamaki Marcin Zukowski Campbell Fraser
Smooth Scan: Statistics-Oblivious Access Paths Renata Borovica-Gajic Stratos Idreos Anastasia Ailamaki Marcin Zukowski Campbell Fraser Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 Q11 Q12 Q13 Q14 Q16 Q18 Q19 Q21 Q22
ZBD: Using Transparent Compression at the Block Level to Increase Storage Space Efficiency
ZBD: Using Transparent Compression at the Block Level to Increase Storage Space Efficiency Thanos Makatos, Yannis Klonatos, Manolis Marazakis, Michail D. Flouris, and Angelos Bilas {mcatos,klonatos,maraz,flouris,bilas}@ics.forth.gr
Jignesh M. Patel. Blog:
Jignesh M. Patel Blog: http://bigfastdata.blogspot.com Go back to the design Query Cache from Processing for Conscious 98s Modern (at Algorithms Hardware least for Hash Joins) 995 24 2 Processor Processor
Using Transparent Compression to Improve SSD-based I/O Caches
Using Transparent Compression to Improve SSD-based I/O Caches Thanos Makatos, Yannis Klonatos, Manolis Marazakis, Michail D. Flouris, and Angelos Bilas {mcatos,klonatos,maraz,flouris,bilas}@ics.forth.gr
Low-Overhead Flash Disaggregation via NVMe-over-Fabrics Vijay Balakrishnan Memory Solutions Lab. Samsung Semiconductor, Inc.
Low-Overhead Flash Disaggregation via NVMe-over-Fabrics Vijay Balakrishnan Memory Solutions Lab. Samsung Semiconductor, Inc. 1 DISCLAIMER This presentation and/or accompanying oral statements by Samsung
TPC-E testing of Microsoft SQL Server 2016 on Dell EMC PowerEdge R830 Server and Dell EMC SC9000 Storage
TPC-E testing of Microsoft SQL Server 2016 on Dell EMC PowerEdge R830 Server and Dell EMC SC9000 Storage Performance Study of Microsoft SQL Server 2016 Dell Engineering February 2017 Table of contents
Low-Overhead Flash Disaggregation via NVMe-over-Fabrics
Low-Overhead Flash Disaggregation via NVMe-over-Fabrics Vijay Balakrishnan Memory Solutions Lab. Samsung Semiconductor, Inc. August 2017 1 DISCLAIMER This presentation and/or accompanying oral statements
Forensic Toolkit System Specifications Guide
Forensic Toolkit System Specifications Guide February 2012 When it comes to performing effective and timely investigations, we recommend examiners take into consideration the demands the software, and
Query processing on raw files. Vítor Uwe Reus
Query processing on raw files Vítor Uwe Reus Outline 1. Introduction 2. Adaptive Indexing 3. Hybrid MapReduce 4. NoDB 5. Summary Outline 1. Introduction 2. Adaptive Indexing 3. Hybrid MapReduce 4. NoDB
CS370 Operating Systems
CS370 Operating Systems Colorado State University Yashwant K Malaiya Fall 2016 Lecture 35 Mass Storage Slides based on Text by Silberschatz, Galvin, Gagne Various sources 1 1 Questions For You Local/Global
COMP283-Lecture 3 Applied Database Management
COMP283-Lecture 3 Applied Database Management Introduction DB Design Continued Disk Sizing Disk Types & Controllers DB Capacity 1 COMP283-Lecture 3 DB Storage: Linear Growth Disk space requirements increases
Hammer Slide: Work- and CPU-efficient Streaming Window Aggregation
Large-Scale Data & Systems Group Hammer Slide: Work- and CPU-efficient Streaming Window Aggregation Georgios Theodorakis, Alexandros Koliousis, Peter Pietzuch, Holger Pirk Large-Scale Data & Systems (LSDS)
Deep Learning Performance and Cost Evaluation
Micron 5210 ION Quad-Level Cell (QLC) SSDs vs 7200 RPM HDDs in Centralized NAS Storage Repositories A Technical White Paper Don Wang, Rene Meyer, Ph.D. info@ AMAX Corporation Publish date: October 25,
Upgrade to Microsoft SQL Server 2016 with Dell EMC Infrastructure
Upgrade to Microsoft SQL Server 2016 with Dell EMC Infrastructure Generational Comparison Study of Microsoft SQL Server Dell Engineering February 2017 Revisions Date Description February 2017 Version 1.0
NVMFS: A New File System Designed Specifically to Take Advantage of Nonvolatile Memory
NVMFS: A New File System Designed Specifically to Take Advantage of Nonvolatile Memory Dhananjoy Das, Sr. Systems Architect SanDisk Corp. 1 Agenda: Applications are KING! Storage landscape (Flash / NVM)
ChunkStash: Speeding Up Storage Deduplication using Flash Memory
ChunkStash: Speeding Up Storage Deduplication using Flash Memory Biplob Debnath +, Sudipta Sengupta *, Jin Li * * Microsoft Research, Redmond (USA) + Univ. of Minnesota, Twin Cities (USA) Deduplication
LATEST INTEL TECHNOLOGIES POWER NEW PERFORMANCE LEVELS ON VMWARE VSAN
LATEST INTEL TECHNOLOGIES POWER NEW PERFORMANCE LEVELS ON VMWARE VSAN Russ Fellows Enabling you to make the best technology decisions November 2017 EXECUTIVE OVERVIEW* The new Intel Xeon Scalable platform
Overview of Data Exploration Techniques. Stratos Idreos, Olga Papaemmanouil, Surajit Chaudhuri
Overview of Data Exploration Techniques Stratos Idreos, Olga Papaemmanouil, Surajit Chaudhuri data exploration not always sure what we are looking for (until we find it) data has always been big volume
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
SoftFlash: Programmable Storage in Future Data Centers Jae Do Researcher, Microsoft Research
SoftFlash: Programmable Storage in Future Data Centers Jae Do Researcher, Microsoft Research 1 The world s most valuable resource Data is everywhere! May. 2017 Values from Data! Need infrastructures for
Accelerating OLTP performance with NVMe SSDs Veronica Lagrange Changho Choi Vijay Balakrishnan
Accelerating OLTP performance with NVMe SSDs Veronica Lagrange Changho Choi Vijay Balakrishnan Agenda OLTP status quo Goal System environments Tuning and optimization MySQL Server results Percona Server
IBM Emulex 16Gb Fibre Channel HBA Evaluation
IBM Emulex 16Gb Fibre Channel HBA Evaluation Evaluation report prepared under contract with Emulex Executive Summary The computing industry is experiencing an increasing demand for storage performance
Hewlett Packard Enterprise HPE GEN10 PERSISTENT MEMORY PERFORMANCE THROUGH PERSISTENCE
Hewlett Packard Enterprise HPE GEN10 PERSISTENT MEMORY PERFORMANCE THROUGH PERSISTENCE Digital transformation is taking place in businesses of all sizes Big Data and Analytics Mobility Internet of Things
Oracle Performance on M5000 with F20 Flash Cache. Benchmark Report September 2011
Oracle Performance on M5000 with F20 Flash Cache Benchmark Report September 2011 Contents 1 About Benchware 2 Flash Cache Technology 3 Storage Performance Tests 4 Conclusion copyright 2011 by benchware.ch
Chapter 12: Query Processing
Chapter 12: Query Processing Database System Concepts, 6 th Ed. See www.db-book.com for conditions on re-use Basic Steps in Query Processing 1. Parsing and translation 2. Optimization 3. Evaluation 12.2
ASN Configuration Best Practices
ASN Configuration Best Practices Managed machine Generally used CPUs and RAM amounts are enough for the managed machine: CPU still allows us to read and write data faster than real IO subsystem allows.
CSCI-GA Database Systems Lecture 8: Physical Schema: Storage
CSCI-GA.2433-001 Database Systems Lecture 8: Physical Schema: Storage Mohamed Zahran (aka Z) mzahran@cs.nyu.edu http://www.mzahran.com View 1 View 2 View 3 Conceptual Schema Physical Schema 1. Create a
BIG DATA AND HADOOP ON THE ZFS STORAGE APPLIANCE
BIG DATA AND HADOOP ON THE ZFS STORAGE APPLIANCE BRETT WENINGER, MANAGING DIRECTOR 10/21/2014 ADURANT APPROACH TO BIG DATA Align to Un/Semi-structured Data Instead of Big Scale out will become Big Greatest
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
CS6453. Data-Intensive Systems: Rachit Agarwal. Technology trends, Emerging challenges & opportuni=es
CS6453 Data-Intensive Systems: Technology trends, Emerging challenges & opportuni=es Rachit Agarwal Slides based on: many many discussions with Ion Stoica, his class, and many industry folks Servers Typical
ADMS/VLDB, August 27 th 2018, Rio de Janeiro, Brazil OPTIMIZING GROUP-BY AND AGGREGATION USING GPU-CPU CO-PROCESSING
ADMS/VLDB, August 27 th 2018, Rio de Janeiro, Brazil 1 OPTIMIZING GROUP-BY AND AGGREGATION USING GPU-CPU CO-PROCESSING OPTIMIZING GROUP-BY AND AGGREGATION USING GPU-CPU CO-PROCESSING MOTIVATION OPTIMIZING
Deep Learning Performance and Cost Evaluation
Micron 5210 ION Quad-Level Cell (QLC) SSDs vs 7200 RPM HDDs in Centralized NAS Storage Repositories A Technical White Paper Rene Meyer, Ph.D. AMAX Corporation Publish date: October 25, 2018 Abstract Introduction
CIS 601 Graduate Seminar. Dr. Sunnie S. Chung Dhruv Patel ( ) Kalpesh Sharma ( )
Guide: CIS 601 Graduate Seminar Presented By: Dr. Sunnie S. Chung Dhruv Patel (2652790) Kalpesh Sharma (2660576) Introduction Background Parallel Data Warehouse (PDW) Hive MongoDB Client-side Shared SQL
SSDs vs HDDs for DBMS by Glen Berseth York University, Toronto
SSDs vs HDDs for DBMS by Glen Berseth York University, Toronto So slow So cheap So heavy So fast So expensive So efficient NAND based flash memory Retains memory without power It works by trapping a small
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
The following is intended to outline our general product direction. It is intended for information purposes only, and may not be incorporated into
The following is intended to outline our general product direction. It is intended for information purposes only, and may not be incorporated into any contract. It is not a commitment to deliver any material,
Ingo Brenckmann Jochen Kirsten Storage Technology Strategists SAS EMEA Copyright 2003, SAS Institute Inc. All rights reserved.
Intelligent Storage Results from real life testing Ingo Brenckmann Jochen Kirsten Storage Technology Strategists SAS EMEA SAS Intelligent Storage components! OLAP Server! Scalable Performance Data Server!
A Comparison of Memory Usage and CPU Utilization in Column-Based Database Architecture vs. Row-Based Database Architecture
A Comparison of Memory Usage and CPU Utilization in Column-Based Database Architecture vs. Row-Based Database Architecture By Gaurav Sheoran 9-Dec-08 Abstract Most of the current enterprise data-warehouses
Analysis of Derby Performance
Analysis of Derby Performance Staff Engineer Olav Sandstå Senior Engineer Dyre Tjeldvoll Sun Microsystems Database Technology Group This is a draft version that is subject to change. The authors can be
Sandor Heman, Niels Nes, Peter Boncz. Dynamic Bandwidth Sharing. Cooperative Scans: Marcin Zukowski. CWI, Amsterdam VLDB 2007.
Cooperative Scans: Dynamic Bandwidth Sharing in a DBMS Marcin Zukowski Sandor Heman, Niels Nes, Peter Boncz CWI, Amsterdam VLDB 2007 Outline Scans in a DBMS Cooperative Scans Benchmarks DSM version VLDB,
Optimizing OLAP Cube Processing on Solid State Drives
Optimizing OLAP Cube Processing on Solid State Drives Zhibo Chen University of Houston Houston, TX 77204, USA Carlos Ordonez University of Houston Houston, TX 77204, USA ABSTRACT Hardware technology has
IBM V7000 Unified R1.4.2 Asynchronous Replication Performance Reference Guide
V7 Unified Asynchronous Replication Performance Reference Guide IBM V7 Unified R1.4.2 Asynchronous Replication Performance Reference Guide Document Version 1. SONAS / V7 Unified Asynchronous Replication
HIGH PERFORMANCE SANLESS CLUSTERING THE POWER OF FUSION-IO THE PROTECTION OF SIOS
HIGH PERFORMANCE SANLESS CLUSTERING THE POWER OF FUSION-IO THE PROTECTION OF SIOS Proven Companies and Products Fusion-io Leader in PCIe enterprise flash platforms Accelerates mission-critical applications
Sorting. Overview. External sorting. Warm up: in memory sorting. Purpose. Overview. Sort benchmarks
15-823 Advanced Topics in Database Systems Performance Sorting Shimin Chen School of Computer Science Carnegie Mellon University 22 March 2001 Sort benchmarks A base case: AlphaSort Improving Sort Performance
Storage. CS 3410 Computer System Organization & Programming
Storage CS 3410 Computer System Organization & Programming These slides are the product of many rounds of teaching CS 3410 by Deniz Altinbuke, Kevin Walsh, and Professors Weatherspoon, Bala, Bracy, and
Building blocks for high performance DWH Computing
Building blocks for high performance DWH Computing Wolfgang Höfer, Nuremberg, 18 st November 2010 Copyright 2010 Fujitsu Technology Solutions Current trends (1) Intel/AMD CPU performance is growing fast
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)
Deduplication Storage System
Deduplication Storage System Kai Li Charles Fitzmorris Professor, Princeton University & Chief Scientist and Co-Founder, Data Domain, Inc. 03/11/09 The World Is Becoming Data-Centric CERN Tier 0 Business
2009. October. Semiconductor Business SAMSUNG Electronics
2009. October Semiconductor Business SAMSUNG Electronics Why SSD performance is faster than HDD? HDD has long latency & late seek time due to mechanical operation SSD does not have both latency and seek
Storage Devices for Database Systems
Storage Devices for Database Systems 5DV120 Database System Principles Umeå University Department of Computing Science Stephen J. Hegner hegner@cs.umu.se http://www.cs.umu.se/~hegner Storage Devices for
REFERENCE ARCHITECTURE Microsoft SQL Server 2016 Data Warehouse Fast Track. FlashStack 70TB Solution with Cisco UCS and Pure Storage FlashArray//X
REFERENCE ARCHITECTURE Microsoft SQL Server 2016 Data Warehouse Fast Track FlashStack 70TB Solution with Cisco UCS and Pure Storage FlashArray//X FLASHSTACK REFERENCE ARCHITECTURE September 2018 TABLE
HyPer-sonic Combined Transaction AND Query Processing
HyPer-sonic Combined Transaction AND Query Processing Thomas Neumann Technische Universität München October 26, 2011 Motivation - OLTP vs. OLAP OLTP and OLAP have very different requirements OLTP high
Azor: Using Two-level Block Selection to Improve SSD-based I/O caches
Azor: Using Two-level Block Selection to Improve SSD-based I/O caches Yannis Klonatos, Thanos Makatos, Manolis Marazakis, Michail D. Flouris, Angelos Bilas {klonatos, makatos, maraz, flouris, bilas}@ics.forth.gr
Quantifying FTK 3.0 Performance with Respect to Hardware Selection
Quantifying FTK 3.0 Performance with Respect to Hardware Selection Background A wide variety of hardware platforms and associated individual component choices exist that can be utilized by the Forensic
Webscale, All Flash, Distributed File Systems. Avraham Meir Elastifile
Webscale, All Flash, Distributed File Systems Avraham Meir Elastifile 1 Outline The way to all FLASH The way to distributed storage Scale-out storage management Conclusion 2 Storage Technology Trend NAND
Introduction to Column Stores with MemSQL. Seminar Database Systems Final presentation, 11. January 2016 by Christian Bisig
Final presentation, 11. January 2016 by Christian Bisig Topics Scope and goals Approaching Column-Stores Introducing MemSQL Benchmark setup & execution Benchmark result & interpretation Conclusion Questions
Speeding Up Cloud/Server Applications Using Flash Memory
Speeding Up Cloud/Server Applications Using Flash Memory Sudipta Sengupta and Jin Li Microsoft Research, Redmond, WA, USA Contains work that is joint with Biplob Debnath (Univ. of Minnesota) Flash Memory
(software agnostic) Computational Considerations
(software agnostic) Computational Considerations The Issues CPU GPU Emerging - FPGA, Phi, Nervana Storage Networking CPU 2 Threads core core Processor/Chip Processor/Chip Computer CPU Threads vs. Cores
Getting the Most Out of SSDs IT System Optimization Best Practices
Getting the Most Out of SSDs IT System Optimization Best Practices Mike Chenery, President and Co-Founder Pliant Technology Storage Myths & Facts Facts: Storage has always been sold on $/GB HDDs are optimized
Overtaking CPU DBMSes with a GPU in Whole-Query Analytic Processing
Overtaking CPU DBMSes with a GPU in Whole-Query Analytic Processing Adnan Agbaria David Minor Natan Peterfreund Eyal Rozenberg Ofer Rosenberg - now at Intel - now at GE Research - now a post-doc at CWI
Anil Vasudeva IMEX Research. (408)
Architecting Next Generation Enterprise Network Storage Anil Vasudeva Research (408) 268-0800 vasudeva@imexresearch.com 2000-2004 Research All rights Reserved Reproduction prohibited, without written permission
I/O. Disclaimer: some slides are adopted from book authors slides with permission 1
I/O Disclaimer: some slides are adopted from book authors slides with permission 1 Thrashing Recap A processes is busy swapping pages in and out Memory-mapped I/O map a file on disk onto the memory space
Data Blocks: Hybrid OLTP and OLAP on compressed storage
Data Blocks: Hybrid OLTP and OLAP on compressed storage Ben Brümmer Technische Universität München Fürstenfeldbruck, 26. November 208 Ben Brümmer 26..8 Lehrstuhl für Datenbanksysteme Problem HDD/Archive/Tape-Storage
Storage Systems Market Analysis Dec 04
Storage Systems Market Analysis Dec 04 Storage Market & Technologies World Wide Disk Storage Systems Market Analysis Wor ldwi d e D i s k Storage S y s tems Revenu e b y Sup p l i e r, 2001-2003 2001
IBM FlashSystem. IBM FLiP Tool Wie viel schneller kann Ihr IBM i Power Server mit IBM FlashSystem 900 / V9000 Storage sein?
FlashSystem Family 2015 IBM FlashSystem IBM FLiP Tool Wie viel schneller kann Ihr IBM i Power Server mit IBM FlashSystem 900 / V9000 Storage sein? PiRT - Power i Round Table 17 Sep. 2015 Daniel Gysin IBM
Introducing Tegile. Company Overview. Product Overview. Solutions & Use Cases. Partnering with Tegile
Tegile Systems 1 Introducing Tegile Company Overview Product Overview Solutions & Use Cases Partnering with Tegile 2 Company Overview Company Overview Te gile - [tey-jile] Tegile = technology + agile Founded
Data Systems that are Easy to Design, Tune and Use. Stratos Idreos
Data Systems that are Easy to Design, Tune and Use data systems that are easy to: (years) (months) design & build set-up & tune (hours/days) use e.g., adapt to new applications, new hardware, spin off
Sun Lustre Storage System Simplifying and Accelerating Lustre Deployments
Sun Lustre Storage System Simplifying and Accelerating Lustre Deployments Torben Kling-Petersen, PhD Presenter s Name Principle Field Title andengineer Division HPC &Cloud LoB SunComputing Microsystems
Column Stores vs. Row Stores How Different Are They Really?
Column Stores vs. Row Stores How Different Are They Really? Daniel J. Abadi (Yale) Samuel R. Madden (MIT) Nabil Hachem (AvantGarde) Presented By : Kanika Nagpal OUTLINE Introduction Motivation Background
Moneta: A High-performance Storage Array Architecture for Nextgeneration, Micro 2010
Moneta: A High-performance Storage Array Architecture for Nextgeneration, Non-volatile Memories Micro 2010 NVM-based SSD NVMs are replacing spinning-disks Performance of disks has lagged NAND flash showed
AN ELASTIC MULTI-CORE ALLOCATION MECHANISM FOR DATABASE SYSTEMS
1 AN ELASTIC MULTI-CORE ALLOCATION MECHANISM FOR DATABASE SYSTEMS SIMONE DOMINICO 1, JORGE A. MEIRA 2, MARCO A. Z. ALVES 1, EDUARDO C. DE ALMEIDA 1 FEDERAL UNIVERSITY OF PARANÁ, BRAZIL 1, UNIVERSITY OF
Toward Seamless Integration of RAID and Flash SSD
Toward Seamless Integration of RAID and Flash SSD Sang-Won Lee Sungkyunkwan Univ., Korea (Joint-Work with Sungup Moon, Bongki Moon, Narinet, and Indilinx) Santa Clara, CA 1 Table of Contents Introduction
Concurrent execution of an analytical workload on a POWER8 server with K40 GPUs A Technology Demonstration
Concurrent execution of an analytical workload on a POWER8 server with K40 GPUs A Technology Demonstration Sina Meraji sinamera@ca.ibm.com Berni Schiefer schiefer@ca.ibm.com Tuesday March 17th at 12:00
Database Acceleration Solution Using FPGAs and Integrated Flash Storage
Database Acceleration Solution Using FPGAs and Integrated Flash Storage HK Verma, Xilinx Inc. August 2017 1 FPGA Analytics in Flash Storage System In-memory or Flash storage based DB reduce disk access
Hardware & System Requirements
Safend Data Protection Suite Hardware & System Requirements System Requirements Hardware & Software Minimum Requirements: Safend Data Protection Agent Requirements Console Safend Data Access Utility Operating
The Benefits of Solid State in Enterprise Storage Systems. David Dale, NetApp
The Benefits of Solid State in Enterprise Storage Systems David Dale, NetApp SNIA Legal Notice The material contained in this tutorial is copyrighted by the SNIA unless otherwise noted. Member companies
FlashStack 70TB Solution with Cisco UCS and Pure Storage FlashArray
REFERENCE ARCHITECTURE Microsoft SQL Server 2016 Data Warehouse Fast Track FlashStack 70TB Solution with Cisco UCS and Pure Storage FlashArray FLASHSTACK REFERENCE ARCHITECTURE December 2017 TABLE OF CONTENTS
Improved Solutions for I/O Provisioning and Application Acceleration
1 Improved Solutions for I/O Provisioning and Application Acceleration August 11, 2015 Jeff Sisilli Sr. Director Product Marketing jsisilli@ddn.com 2 Why Burst Buffer? The Supercomputing Tug-of-War A supercomputer
Chapter 10: Mass-Storage Systems. Operating System Concepts 9 th Edition
Chapter 10: Mass-Storage Systems Silberschatz, Galvin and Gagne 2013 Chapter 10: Mass-Storage Systems Overview of Mass Storage Structure Disk Structure Disk Attachment Disk Scheduling Disk Management Swap-Space
Persistent Memory in Mission-Critical Architecture (How and Why) Adam Roberts Engineering Fellow, Western Digital Corporation
Persistent Memory in Mission-Critical Architecture (How and Why) Adam Roberts Engineering Fellow, Western Digital Corporation Forward-Looking Statements Safe Harbor Disclaimers This presentation contains
Performance Testing of SQL Server on Kaminario K2 Storage
Performance Testing of SQL Server on Kaminario K2 Storage September 2016 TABLE OF CONTENTS 2 3 5 14 15 17 Executive Summary Introduction to Kaminario K2 Performance Tests for SQL Server Summary Appendix:
Discover the all-new CacheMount
Discover the all-new CacheMount 1 2 3 4 5 Why CacheMount and what are its problem solving abilities Cache function makes the hybrid cloud more efficient The key of CacheMount: Cache Volume User manual
Distributed File Systems Part IV. Hierarchical Mass Storage Systems
Distributed File Systems Part IV Daniel A. Menascé Hierarchical Mass Storage Systems On-line data requirements Mass Storage Systems Concepts Mass storage system architectures Example systems Performance
ZEST Snapshot Service. A Highly Parallel Production File System by the PSC Advanced Systems Group Pittsburgh Supercomputing Center 1
ZEST Snapshot Service A Highly Parallel Production File System by the PSC Advanced Systems Group Pittsburgh Supercomputing Center 1 Design Motivation To optimize science utilization of the machine Maximize
Highly Scalable, Non-RDMA NVMe Fabric. Bob Hansen,, VP System Architecture
A Cost Effective,, High g Performance,, Highly Scalable, Non-RDMA NVMe Fabric Bob Hansen,, VP System Architecture bob@apeirondata.com Storage Developers Conference, September 2015 Agenda 3 rd Platform
Moneta: A High-Performance Storage Architecture for Next-generation, Non-volatile Memories
Moneta: A High-Performance Storage Architecture for Next-generation, Non-volatile Memories Adrian M. Caulfield Arup De, Joel Coburn, Todor I. Mollov, Rajesh K. Gupta, Steven Swanson Non-Volatile Systems
Bring x3 Spark Performance Improvement with PCIe SSD. Yucai, Yu BDT/STO/SSG January, 2016
Bring x3 Spark Performance Improvement with PCIe SSD Yucai, Yu (yucai.yu@intel.com) BDT/STO/SSG January, 2016 About me/us Me: Spark contributor, previous on virtualization, storage, mobile/iot OS. Intel
HyPer-sonic Combined Transaction AND Query Processing
HyPer-sonic Combined Transaction AND Query Processing Thomas Neumann Technische Universität München December 2, 2011 Motivation There are different scenarios for database usage: OLTP: Online Transaction
Decompressing Snappy Compressed Files at the Speed of OpenCAPI. Speaker: Jian Fang TU Delft
Decompressing Snappy Compressed Files at the Speed of OpenCAPI Speaker: Jian Fang TU Delft 1 Current Project SHADE Scalable Heterogeneous Accelerated DatabasE Spark DB CPU POWER9 ARROW DNA Seq Sort Join
Computer Systems Laboratory Sungkyunkwan University
I/O System Jin-Soo Kim (jinsookim@skku.edu) Computer Systems Laboratory Sungkyunkwan University http://csl.skku.edu Introduction (1) I/O devices can be characterized by Behavior: input, output, storage
Time-Series Data in MongoDB on a Budget. Peter Schwaller Senior Director Server Engineering, Percona Santa Clara, California April 23th 25th, 2018
Time-Series Data in MongoDB on a Budget Peter Schwaller Senior Director Server Engineering, Percona Santa Clara, California April 23th 25th, 2018 TIME SERIES DATA in MongoDB on a Budget Click to add text
Single-pass restore after a media failure. Caetano Sauer, Goetz Graefe, Theo Härder
Single-pass restore after a media failure Caetano Sauer, Goetz Graefe, Theo Härder 20% of drives fail after 4 years High failure rate on first year (factory defects) Expectation of 50% for 6 years https://www.backblaze.com/blog/how-long-do-disk-drives-last/