S1: Csci 8701 - Meeting 1
Copyright: S. Shekhar, C. S. Dept., University of Minnesota, Minneapolis, MN 55455. ,,.(home)
S2: Administrative
  • Introductions
    • Instructor
    • TAs
    • Peers
  • Syllabus
    • Schedules: lectures, homeworks, exams, recitations
    • Textbooks, Reference material
    • Topics
  • Policies
    • Late homeworks
    • Cheating
    • Waiting List - Attendance*, estimates


Copyright: S. Shekhar, C. S. Dept., University of Minnesota, Minneapolis, MN 55455. ,,.(home)
S3: Goals
  • Understand Database Research
    • Background, Scope
    • Accomplishments of last 2/3 decades
    • Next Challenges
  • Course themes
    • Research Topics in DB Systems
    • Common Methodologies
    • Common platforms and applications
  • Out of Scope
    • Database Theory - e.g. Deductive databases


Copyright: S. Shekhar, C. S. Dept., University of Minnesota, Minneapolis, MN 55455. ,,.(home)
S4: Overview
  • NSF Workshop (2/1990)
    • Goal: Technology pull factors
    • for adv. DB technology
  • Main predictions
    • Many new pull technologies
    • Next gen. DB >> Business DBMSs
    • Inter-organization databases
  • Database challenges
    • type extensions, complex objects
    • rules, multimedia, archival
    • large-scale, heterogeneous, distributed


Copyright: S. Shekhar, C. S. Dept., University of Minnesota, Minneapolis, MN 55455. ,,.(home)
S5: Discussion Questions
  • Justify the need for DB research when
    • someone claims that
    • (i) Main memory is cheaper and bigger
    • (2) CPUs and Networks are faster
    • (3) DB technology is mature and
    • Commercial DBMSs are commodities


Copyright: S. Shekhar, C. S. Dept., University of Minnesota, Minneapolis, MN 55455. ,,.(home)
S6: Background, Scope
  • History
    • 1960s - mostly industry
    • 1970s & 80s- universities, federal labs
  • Applications
    • Business applications
    • Banking, record keeping, reservation systems
  • Requirements
    • Efficiency
    • Resilience
    • Access Control
    • Persistence


Copyright: S. Shekhar, C. S. Dept., University of Minnesota, Minneapolis, MN 55455. ,,.(home)
S7: Accomplishments of 70s, 80s
  • Three Areas
    • Relational Databases
    • Transaction Management
    • Distributed Databases
  • Relational Databases
    • Declarative Query Language
    • Query optimization - theory, algos.
    • DB Design: Normalization
    • Page level Record clustering algo
    • Buffer Management algo.
    • Indexing techniques
    • Implemented prototype relation DBMSs


Copyright: S. Shekhar, C. S. Dept., University of Minnesota, Minneapolis, MN 55455. ,,.(home)
S8: Accomplishments of 70s, 80s
  • Transaction Management
    • Transaction concept - ACID properties
    • Serializability theory
    • Concurrency control - 2PL, TSO, versions
    • Recovery Algos: Write ahead logging
    • Recovery Algos: Shadow paging
    • Recovery Algos: Backup on separate device
  • Distributed Databases
    • communication cost, location transparency
    • Improving availability
    • Algos for QO, CC, Recovery, Copy mgmt
    • CC : two phase commit


Copyright: S. Shekhar, C. S. Dept., University of Minnesota, Minneapolis, MN 55455. ,,.(home)
S9: Discussion Questions
  • Review the topics in weekly schedule
    • Match them w/ accomplishments of 70s, 80s
    • Do we cover topics outside accomplishments list?
  • List validation methodologies in DB research.
  • Compare internet with distributed databases
    • Goals, platform, applications
    • Query Language
  • Authors did not list the following:
    • DB design: Entity Relationship Model
    • Would you consider it a major accomplishment? Why?
    • What else would you put on list of accomplishments?


Copyright: S. Shekhar, C. S. Dept., University of Minnesota, Minneapolis, MN 55455. ,,.(home)
S10: The Nest Challenges
  • Application Domains
    • NASA EOS
    • Computer Aided Design
    • DNA Sequencing
    • Mining Buying Patterns
    • Insurance - multimedia
  • Issues
    • New complex data types
    • (Storage, querying, Type Management)
    • Rule Processing
    • Modeling Spatial and Temporal data
    • Scaling Up, Parallelism
    • Tertiary Storage, Long term xactions
    • Versions, Configurations
    • Heterogeneous, Distributed Databases
    • Mediator, Security, ...


Copyright: S. Shekhar, C. S. Dept., University of Minnesota, Minneapolis, MN 55455. ,,.(home)
S11: Discussion Questions
  • Review the topics in weekly schedule
    • Match them w/ next challenges for 90s
    • Do we cover any of these ?
  • Crystal ball evaluation:
    • Q? Name major info. technologies of 1990s.
    • Q? Name major killer applications of 1990s.
    • Q? Which were not anticipated by authors?
    • Q? Which challenge was paramount in last decade?
  • Redo previous question for the coming decade.
  • Hints: Internet, EC, CRM, DWH, DM, Wireless


Copyright: S. Shekhar, C. S. Dept., University of Minnesota, Minneapolis, MN 55455. ,,.(home)
S12: Discussion Questions
  • Find report of a recent workshop (e.g. Asilomar)
    • on future directions for Database Research.
    • Compare the challenges with those in the paper.
  • Review the Challenges.
    • Which are these relevant to internet?
    • How are they handled on internet?
    • Are these still research challenges?


Copyright: S. Shekhar, C. S. Dept., University of Minnesota, Minneapolis, MN 55455. ,,.(home)