IDISK Cluster 8 disks, 8 CPUs, DRAM /shelf

Slides:



Advertisements
Similar presentations
Exadata Distinctives Brown Bag New features for tuning Oracle database applications.
Advertisements

Daniel Schall, Volker Höfner, Prof. Dr. Theo Härder TU Kaiserslautern.
6.830 Lecture 9 10/1/2014 Join Algorithms. Database Internals Outline Front End Admission Control Connection Management (sql) Parser (parse tree) Rewriter.
Parallel Databases Michael French, Spencer Steele, Jill Rochelle When Parallel Lines Meet by Ken Rudin (BYTE, May 98)
High Performance Analytical Appliance MPP Database Server Platform for high performance Prebuilt appliance with HW & SW included and optimally configured.
1. Aim High with Oracle Real World Performance Andrew Holdsworth Director Real World Performance Group Server Technologies.
External Sorting CS634 Lecture 10, Mar 5, 2014 Slides based on “Database Management Systems” 3 rd ed, Ramakrishnan and Gehrke.
Voyager Interest Group Voyager Access Reports: what they are and how they work October 29, 2008.
IBM RS6000/SP Overview Advanced IBM Unix computers series Multiple different configurations Available from entry level to high-end machines. POWER (1,2,3,4)
1 Hash-Join n Hybrid hash join m R: 71k rows x 145 B m S: 200k rows x 165 B m TPC-D lineitem, part n Clusters benefit from one-pass algorithms n IDISK.
®  SAP AG 1998 Memory Based Computing (Gemma F. Durany) / 1 Gemma F. Durany SAP AG Memory Based Computing.
VLDB Revisiting Pipelined Parallelism in Multi-Join Query Processing Bin Liu and Elke A. Rundensteiner Worcester Polytechnic Institute
Server Platforms Week 11- Lecture 1. Server Market $ 46,100,000,000 ($ 46.1 Billion) Gartner.
1 External Sorting for Query Processing Yanlei Diao UMass Amherst Feb 27, 2007 Slides Courtesy of R. Ramakrishnan and J. Gehrke.
Fall 2008Parallel Databases1. Fall 2008Parallel Databases2 Ideal Parallel Systems Two key properties:  Linear Speedup: Twice as much hardware can perform.
5 Creating the Physical Model. Designing the Physical Model Phase IV: Defining the physical model.
1 A Comparison of Approaches to Large-Scale Data Analysis Pavlo, Paulson, Rasin, Abadi, DeWitt, Madden, Stonebraker, SIGMOD’09 Shimin Chen Big data reading.
Copyright © 2013, Oracle and/or its affiliates. All rights reserved. 1 Preview of Oracle Database 12 c In-Memory Option Thomas Kyte
Parallel Execution Plans Joe Chang
SQL Server 2005 Performance Enhancements for Large Queries Joe Chang
Analyzing the Energy Efficiency of a Database Server Hanskamal Patel SE 521.
Shilpa Seth.  Centralized System Centralized System  Client Server System Client Server System  Parallel System Parallel System.
Ling Guo Feb 15, 2010 Database(RDBMS) Software Review Oracle RDBMS (Oracle Cooperation) 4()6 Oracle 10g Express version DB2 (IBM) IBM DB2 Express-C SQL.
Parallel DBMS Instructor : Marina Gavrilova
PHP Data Object (PDO) Khaled Al-Sham’aa. What is PDO? PDO is a PHP extension to formalise PHP's database connections by creating a uniform interface.
1 © Prentice Hall, 2002 Physical Database Design Dr. Bijoy Bordoloi.
Oracle Challenges Parallelism Limitations Parallelism is the ability for a single query to be run across multiple processors or servers. Large queries.
Data Warehousing 1 Lecture-24 Need for Speed: Parallelism Virtual University of Pakistan Ahsan Abdullah Assoc. Prof. & Head Center for Agro-Informatics.
Databases - their construction and use John Dubery, U3A Intermediate Computing April 2011.
Parallel Database Systems Instructor: Dr. Yingshu Li Student: Chunyu Ai.
IBM xSeries - Linux Installation/Performance Optimization Thousands of IT Professionals before you have already passed their certification.
1 Tandem Daytona TeraByte Sort: Tsort 1 TB in 47.5 Minutes Daivd Cossock, Sam Fineberg, Pankaj Mehra, John Peck Trophy presentation by Jim Gray.
EEL 5708 Cluster computers. Case study: Google Lotzi Bölöni.
PHP Workshop ‹#› PHP Data Object (PDO). PHP Workshop ‹#› What is PDO? PDO is a PHP extension to formalise PHP's database connections by creating a uniform.
Infrastructure for Data Warehouses. Basics Of Data Access Data Store Machine Memory Buffer Memory Cache Data Store Buffer Bus Structure.
Mapping the Data Warehouse to a Multiprocessor Architecture
Introduction to Database Systems1 External Sorting Query Processing: Topic 0.
Database Management Systems 3ed, R. Ramakrishnan and J. Gehrke1 External Sorting Chapters 13: 13.1—13.5.
Database Management Systems, R. Ramakrishnan and J. Gehrke1 External Sorting Chapter 11.
Handling Data Skew in Parallel Joins in Shared-Nothing Systems Yu Xu, Pekka Kostamaa, XinZhou (Teradata) Liang Chen (University of California) SIGMOD’08.
How to kill SQL Server Performance Håkan Winther.
What Should a DBMS Do? Store large amounts of data Process queries efficiently Allow multiple users to access the database concurrently and safely. Provide.
Exadata Distinctives 988 Bobby Durrett US Foods. What is Exadata? Complete Oracle database platform Disk storage system Unique to Exadata – intelligent.
Introduction to Parallel Computing: MPI, OpenMP and Hybrid Programming
Hardware Technology Trends and Database Opportunities
Chapter 17: Database System Architectures
Operating System.
Lecture 16: Data Storage Wednesday, November 6, 2006.
Berkeley Cluster: Zoom Project
Oracle Storage Performance Studies
Mapping the Data Warehouse to a Multiprocessor Architecture
April 30th – Scheduling / parallel
Chapter 17: Database System Architectures
External Sorting The slides for this text are organized into chapters. This lecture covers Chapter 11. Chapter 1: Introduction to Database Systems Chapter.
Akshay Tomar Prateek Singh Lohchubh
External Joins Query Optimization 10/4/2017
Optifacts Enhanced Reporting Application
Selected Topics: External Sorting, Join Algorithms, …
Parallel Analytic Systems
Query Optimization CS 157B Ch. 14 Mien Siao.
The PROCESS of Queries John Deardurff Website: ThatAwesomeTrainer.com
Introduction to Teradata
Parallel DBMS Chapter 22, Part A
Parallel DBMS Chapter 22, Sections 22.1–22.6
Performance And Scalability In Oracle9i And SQL Server 2000
Database System Architectures
Parallel DBMS DBMS Textbook Chapter 22
IRAM Vision Microprocessor & DRAM on a single chip:
External Sorting Dina Said
Cluster Computers.
Presentation transcript:

IDISK Cluster 8 disks, 8 CPUs, DRAM /shelf 15 shelves /rack = 120 disks/rack 1312 disks / 120 = 11 racks Connect 4 disks / ring 1312 / 4 = 328 1.5 Gbit links 328 / 16 => 36 32x32 switch HW, assembly cost: ~$1.5 M

Cluster IDISK Software Models 1) Shared Nothing Database: (e.g., IBM, Informix, NCR TeraData, Tandem) 2) Hybrid SMP Database: Front end running query optimizer, applets downloaded into IDISKs 3) Start with Personal Database code developed for portable PCs, PDAs (e.g., M/S Access, M/S SQLserver, Oracle Lite, Sybase SQL Anywhere) then augment with new communication software

Back of the Envelope Benchmarks All configurations have ~300 disks Equivalent speeds for central and disk procs. Benchmarks: Scan, Sort, Hash-Join

Scan Scan 6 billion 145 B rows TPC-D lineitem table Embarrassingly parallel task; limited by number processors IDISK Speedup: NCR: 2.4X Compaq: 12.6X 12.6X 2.4X

MinuteSort External sorting: data starts and ends on disk MinuteSort: how much can we sort in a minute? Benchmark designed by Nyberg, et al., SIGMOD ‘94 Current record: 8.4 GB on 95 UltraSPARC I’s w/ Myrinet [NOWSort:Arpaci-Dusseau97] Sorting Review: One-pass sort: data sorted = memory size Two-pass sort: Data sorted proportional to sq.rt. (memory size) Disk I/O requirements: 2x that of one-pass sort

MinuteSort IDISK sorts 2.5X - 13X more than clusters IDISK sort limited by disk B/W Cluster sorts limited by network B/W