1 Rethinking Database System Architecture: Towards a Self-Tuning RISC-style Database System Surajit Chaudhuri Gerhard Weikum Microsoft Research University.

Slides:



Advertisements
Similar presentations
Tales from the Lab: Experiences and Methodology Demand Technology User Group December 5, 2005 Ellen Friedman SRM Associates, Ltd.
Advertisements

1 Perspectives from Operating a Large Scale Website Dennis Lee VP Technical Operations, Marchex.
A Ridiculously Easy & Seriously Powerful SQL Cloud Database Itamar Haber AVP Ops & Solutions.
January 30, 2014 Copyright Jim Farley Beyond JDBC: Java Object- Relational Mappings Jim Farley e-Commerce Program Manager GE Research and Development
Hello i am so and so, title/role and a little background on myself (i.e. former microsoft employee or anything interesting) set context for what going.
Chapter 9. Performance Management Enterprise wide endeavor Research and ascertain all performance problems – not just DBMS Five factors influence DB performance.
Module 13: Performance Tuning. Overview Performance tuning methodologies Instance level Database level Application level Overview of tools and techniques.
2 A bank application needs to access information from the customer database and integrate it with loan credit history information stored in a legacy database.
Database Architectures and the Web
SSRS 2008 Architecture Improvements Scale-out SSRS 2008 Report Engine Scalability Improvements.
Multi-Mode Survey Management An Approach to Addressing its Challenges
Agile Infrastructure built on OpenStack Building The Next Generation Data Center with OpenStack John Griffith, Senior Software Engineer,
Emre Yenier Rethinking Database System Architecture: Towards a Self-tuning RISC-style Database System.
5.1 © 2007 by Prentice Hall 5 Chapter Foundations of Business Intelligence: Databases and Information Management.
VoipNow Core Solution capabilities and business value.
UNIT-e Research & Development Microsoft Technology Day Stephen Cain (System Architect)
Network Management Overview IACT 918 July 2004 Gene Awyzio SITACS University of Wollongong.
Technical Architectures
Integration and Insight Aren’t Simple Enough Laura Haas IBM Distinguished Engineer Director, Computer Science Almaden Research Center.
Architectural Design Principles. Outline  Architectural level of design The design of the system in terms of components and connectors and their arrangements.
Chapter 4: Database Management. Databases Before the Use of Computers Data kept in books, ledgers, card files, folders, and file cabinets Long response.
© 2014 ScaleArc. All Rights Reserved. 1 Creating an Agile Data Environment for Apps in the Cloud Summer 2014.
 MODERN DATABASE MANAGEMENT SYSTEMS OVERVIEW BY ENGINEER BILAL AHMAD
Sitefinity Performance and Architecture
1DBTest2008. Motivation Background Relational Data Warehousing (DW) SQL Server 2008 Starjoin improvement Testing Challenge Extending Enterprise-class.
1 Overview of Database Federation and IBM Garlic Project Presented by Xiaofen He.
Oracle10g for Data Warehousing Jiangang Luo
Query Optimization Allison Griffin. Importance of Optimization Time is money Queries are faster Helps everyone who uses the server Solution to speed lies.
Architecture of the R/3 System Chapter 14 C & L Chapter 8 M & W.
Context Tailoring the DBMS –To support particular applications Beyond alphanumerical data Beyond retrieve + process –To support particular hardware New.
AL-MAAREFA COLLEGE FOR SCIENCE AND TECHNOLOGY INFO 232: DATABASE SYSTEMS CHAPTER 1 DATABASE SYSTEMS (Cont’d) Instructor Ms. Arwa Binsaleh.
1 Introduction to Database Systems. 2 Database and Database System / A database is a shared collection of logically related data designed to meet the.
Oracle Challenges Parallelism Limitations Parallelism is the ability for a single query to be run across multiple processors or servers. Large queries.
Data Warehouse Overview September 28, 2012 presented by Terry Bilskie.
© 2007 by Prentice Hall 1 Introduction to databases.
Towards a Grid-based DBMS Craig Thompson University of Arkansas In certain high-end data-centric applications, practitioners are discovering that traditional.
Engr. M. Fahad Khan Lecturer Software Engineering Department University Of Engineering & Technology Taxila.
Component Technology. Challenges Facing the Software Industry Today’s applications are large & complex – time consuming to develop, difficult and costly.
Database Design and Management CPTG /23/2015Chapter 12 of 38 Functions of a Database Store data Store data School: student records, class schedules,
Chapter 18 Object Database Management Systems. McGraw-Hill/Irwin © 2004 The McGraw-Hill Companies, Inc. All rights reserved. Outline Motivation for object.
Views Lesson 7.
1 © 1999 Microsoft Corp.. Microsoft Repository Phil Bernstein Microsoft Corp.
MANAGING DATA RESOURCES ~ pertemuan 7 ~ Oleh: Ir. Abdul Hayat, MTI.
Distributed Information Systems. Motivation ● To understand the problems that Web services try to solve it is helpful to understand how distributed information.
1 CMPT 275 High Level Design Phase Modularization.
DEV14 – Building Business Dashboards: Excel Services, KPIs and Report Centers Darwin Schweitzer Enterprise Technology Strategist
Creating SmartArt 1.Create a slide and select Insert > SmartArt. 2.Choose a SmartArt design and type your text. (Choose any format to start. You can change.
Review of Parnas’ Criteria for Decomposing Systems into Modules Zheng Wang, Yuan Zhang Michigan State University 04/19/2002.
Object storage and object interoperability
Chapter 18 Object Database Management Systems. Outline Motivation for object database management Object-oriented principles Architectures for object database.
Getting the Architecture Right Jeffrey D. Taft, PhD Chief Architect for Electric Grid Transformation Pacific Northwest National Laboratory March 17, 2016.
DATABASE PERFORMANCE COOKBOOK.  Great news!  You’ve completed CSCI 6442 (congratulations!)  You have an MS from GWU (congratulations again!)  You’re.
SAP BI – The Solution at a Glance : SAP Business Intelligence is an enterprise-class, complete, open and integrated solution.
1 Copyright © 2008, Oracle. All rights reserved. Repository Basics.
BIG DATA/ Hadoop Interview Questions.
Managing Data Resources File Organization and databases for business information systems.
Journey to the HyperConverged Agile Infrastructure
Integrating Enterprise Applications Into SharePoint® Portal Server
N-Tier Architecture.
Query Optimization for Object-Relational Database Systems
Migrating Your BI Platform To Azure
Database Management System (DBMS)
Data Warehouse Overview September 28, 2012 presented by Terry Bilskie
Evaluating Transaction System Performance
Cloud Data Services: Self-Manageability and other Challenges
LitwareHR v2: an S+S reference application
UNIT 5 EMBEDDED SYSTEM DEVELOPMENT
UNIT 5 EMBEDDED SYSTEM DEVELOPMENT
Andy Puckett – Sales Engineer
Games Development 2 Entity / Architecture Review
Presentation transcript:

1 Rethinking Database System Architecture: Towards a Self-Tuning RISC-style Database System Surajit Chaudhuri Gerhard Weikum Microsoft Research University of the Saarland Redmond, USA Saarbruecken, Germany

2 Conclusion Problem: DBMS technology is packaged monolithically too many features, too much complexity Solution: RISC-style simplification and componentization  break up DBMS into layered packages with narrow APIs and self-tuning capabilities  compose appropriate packages into broader range of IT applications Think globally, fix locally

3 Outline Analysis Role Models for New Departure Proposal

4 Passing of a Dream Old WorldNew World DBMS DBMS at center of the universe payroll inventory order entry ERP dot com Web server Mining multi-tier architecture with many custom „data managers“

5 Why Did This Happen? Universality of DBMS was a leap of faith SQL is unnatural and complex –Yet another failed example of transparency trap Featurism has turned into a curse –Excessive bundling –Performance is unpredictable –(Auto-) Tuning is a nightmare Unacceptable GPR for app system architects

6 Example of Poor GPR: DBMS Query Processor Yet another indexing smart added Yet another join method added Yet another transformation rule added Optimizer designers will admit –It is unpredictable –Hard to abstract principles ERP/Mining/etc attempt to outsmart QP Turning into black magic –Cannot educate next generation of engineers

7 Role Models for New Departure Ex. 1: Aircraft with many subsystems (engine, fuselage, electrical control, etc.) Ex. 2: RISC hardware No single engineer understands entire system Local theories for individual subsystems and reasonable understanding of interactions –Few points of interaction with stable and narrow interfaces –Built-in system support for debugging subcomponents (incl. performance)

8 RISC Philosophy for DBMS DBMS technology must be packaged as components with simplified functionality Enforce –Layered approach –Strong limits on interaction (narrow APIs) –Multiple consumers for a component Components must have manageable complexity to be desirable for its potential consumers Encapsulation must include predictable performance and self-tuning

9 Why Predictability is Crucial From best-effort to guaranteed performance ”Our ability to analyze and predict the performance of the enormously complex software systems... are painfully inadequate" (PITAC Report) Downtime is very expensive (100K$/min) Very slow servers are like unavailable servers Tuning for peak load requires predictability of workload  config  performance function Self-tuning requires mathematical models – Feasible at component scale

10 Internal Server Error. Our system administrator has been notified. Please try later again. Check Availability (Look-Up Will Take 8-25 Seconds)

11 RISC-style Engine (Components) Level 1 (base layer): SPJ only –only B-trees, with automatic index selection built-in –API includes prioritization & exec. time prediction Level 2: Support for aggregation –Uses level 1 with narrow API –Self-tuning for aggregation considerations Level 3: Full-fledged SQL Layering sacrifices performance for manageability Design principles for components: include only functionality that is self-tuning apply Occam‘s razor for internal alternatives

12 RISC in the Large Use level 1 engine (SPJ, or merely record and index managers) for MP3 repository, simple E-service etc. Use level 2 engine (SPJ + aggregation) for OLAP or ERP Use level 3 engine (SQL) for full-fledged DW, legacy apps Composition principles for IT solutions in the large: Choose least-complexity components IT solution can rely on predictable/guaranteed performance of components

13 Implications of RISC Approach Need for Universal Glue for components –COM/Universal Runtime and EJB Simplicity is key –Eliminate all second-order optimizations Restrict alternatives –Not yet another join method or transformation rule –Don’t abuse extensibility!

14 Road Map Demonstrate “plug and play” light-weight data servers for various scenarios (API and guaranteed performance): – MP3 repositories – OLAP server – Metadata manager Open source “bazaar”?

15 Potential Caveats and Rebuttals We’ve been down this road before! But we now have better understanding of the appropriate components and APIs. We will lose performance! But we win in terms of predictability and overall GPR. There is no business incentive! As industries mature, predictability and manageability do matter for long-term benefit.