Transcript DBMS Database System

ECE 569
Database System Engineering
Spring 2004
Yanyong Zhang www.ece.rutgers.edu/~yyzhang
Course URL www.ece.rutgers.edu/~yyzhang/spring04
ECE 569 Database System Engineering
Spring 2004
About the instructor (Yanyong Zhang)
Yanyong
 Office: Core 518
 Office hours: TBD (usually before class)
 Office number: 5-0608
 Email: [email protected]
 URL: www.ece.rutgers.edu/~yyzhang
 Research interests:

● distributed computing
operating systems
● sensor networks
●
ECE 569 Database System Engineering
Spring 2004
Something about the background

What is database?

a very large, integrated collection of data

Query

Transaction


A group of queries which possess the ACID (atomic,
consistent, isolated, and durable) property
DBMS (DataBase Management System)

a software package designed to store and manage
databases
ECE 569 Database System Engineering
Spring 2004
Overview
User programs
Database
System
Application programs / Queries
DBMS
Software to process queries
Software to access stored data
Stored
database
definition
ECE 569 Database System Engineering
Stored
database
Spring 2004
DBMS Overview

A database management system (DBMS) provides
efficient access to large amounts of persistent
data

Data models and query languages allow efficient access
while hiding complexity from users

Efficient shared access requires concurrency.
Transactions provide transparency to this concurrency.
Application programs are easier to write.

In many cases the data is valuable. It must be protected
from the effects of failure (resiliency) and sabotage
(security).
ECE 569 Database System Engineering
Spring 2004
Files vs. DBMS

Applications must stage large datasets between
main memory and secondary storage (e.g.,
buffering, page-oriented access, 32-bit addressing)

Special code for different queries

Must protect data from inconsistency due to
multiple concurrent users

Crash recovery

Security and access control
ECE 569 Database System Engineering
Spring 2004
Why DBMS?

Data independence and efficient data access

Reduced application development time

Data integrity and security

Uniform data administration

Concurrent accesses, recovery from crashes
ECE 569 Database System Engineering
Spring 2004
Data Models

A data model is a collection of concepts for
describing data

A schema is a description of a particular collection
of data, using a given data model.

The relational model of data is the most widely
used model today

Main concept: relation, basically a table with rows and
columns

Every relation has a schema, which describes the
columns, or fields
ECE 569 Database System Engineering
Spring 2004
Levels of Abstractions

Abstraction is used to hide complexity and allow for a separation of
concerns (What vs. How).

Many views, single conceptual (logical) schema, and single
physical schema

Views describe how users see the data

Conceptual schema defines logical structure

Physical schema describes the files and indexes used.
View 1
Subschema
definition
language
Data
definition
language
ECE 569 Database System Engineering
View 2
View 3
Conceptual Schema
Physical Schema
Specialized view
of enterprise
Complete model
of enterprise
Records,
pointers, indices
Spring 2004
Example


Sample applications

Admit_patient

Make_diagnosis

Record_vital_signs
In relational data model we can express schema with
following tables:

patient (name, address, balance_due, room#)

payments (name, amount, date)

vital_signs (name, pulse, bp, time)

diagnosis (patient_name, disease_name)

disease (disease_name, treatment)
ECE 569 Database System Engineering
Spring 2004
Examples

Physical Level

Specify indices, e.g.,
CREATE INDEX room_index ON patient(room#);


Specify performance related characteristics of relations
Conceptual Level

Define tables, specifying data types for each attribute.
CR
CREATE TABLE patient (
name char(30),
address char(100),
balance_due number(6,2),
room# integer,
PRIMARY KEY (name));
ECE 569 Database System Engineering
Spring 2004
Examples – cont’d

External Level

Define views for various purposes, e.g.,
CREATE VIEW doctor-view-diagnosis AS
SELECT name, room#, disease_name,treatment
FROM patient, diagnosis, diseases
WHERE name = patient_name AND
diagnosis.disease_name = disease.disease_name;
ECE 569 Database System Engineering
Spring 2004
Data Independence

Applications insulated from how data is structured
and stored

Logical data independence: protection from
change in logical structure of data

Physical data independence: protection from
changes in physical structure of data
ECE 569 Database System Engineering
Spring 2004
Concurrency Control

Concurrent execution of user programs is
essential for good DBMS performance

Why??

Interleaving actions of different user programs can
lead to inconsistency: e.g., check is cleared while
account balance is being computed

DBMS ensures such problems don’t arise: users
can pretend they are using a single-user system
ECE 569 Database System Engineering
Spring 2004
Transaction: An execution of a DB program

Key concept is transaction, which is an atomic
sequence of database actions

Each transaction, executed completely, must leave
the DB in a consistent state if DB is consistent
when the transaction begins.

Users can specify some simple integrity constraints on
the data, and DBMS will enforce them

DBMS doesn’t understand the semantics of the data

Ensuring that a transaction (run alone) preserves
consistency is ultimately the user’s responsibility.
ECE 569 Database System Engineering
Spring 2004
Scheduling concurrent transactions

DBMS ensures that execution of {T1, T2, …,
equivalent to some serial execution T1’…Tn’.


Tn}
is
locking scheme
Two-phase locking
ECE 569 Database System Engineering
Spring 2004
Ensuring atomicity

DBMS ensures atomicity (all-or-nothing property)
even if system crashes in the middle of a Xact.

Idea: keep a log (history) of all actions carried out
by the DBMS while executing a set of Xacts
ECE 569 Database System Engineering
Spring 2004
Structure of a DBMS

A typical DBMS has a
layered architecture

The figure does not
show the concurrency
control and recovery
component

This is one of several
possible architectures;
each system has its
own variations.
ECE 569 Database System Engineering
Query optimization and
execution
Relational operators
Files and access methods
Buffer management
Disk space management
Spring 2004
About the course



What will we focus on?

Relational data model

Transaction processing

DBMS design
What will we not focus on?

OO data model, etc

SQL programming
Goal

Understand DBMS design issues

Develop background for research in database area
ECE 569 Database System Engineering
Spring 2004
What should you’ve know

Data structure and algorithms

Operations system knowledge

C, Unix

Background in data model and query languages
recommended
ECE 569 Database System Engineering
Spring 2004
What will you encounter - topics

1. Relational Data Model (2-4)

2. DBMS Design / Implementation (5-11)


a) File organization (5-6)

b) Access methods (7-9)

c) Query processing (10-11)
3. Transaction Processing

a) Transaction Models (12-13)

b) Isolation (14-20)

c) Performance (21-22)

d) B-tree Synchronization (23-24)

e) Recovery (25-29)
ECE 569 Database System Engineering
Spring 2004
What will you encounter - projects

Projects

Develop a client/server relational DBMS
- Query processing / Physical data model / Data dictionary
- Concurrency control / Recovery

Work in groups of at most 4.
- You may choose groups but I must approve.
- At least three members of each group should be strong C
programmers.

Projects are difficult and time-consuming.
- ~10K lines of codes
- Use threads and RPC
- Code is difficult to debug

Projects are interesting and rewarding.
ECE 569 Database System Engineering
Spring 2004
Grading Policy

3 Homework assignments (15%)

Project (45%)

Two exams (20% each)

Course URL:
www.ece.rutgers.edu/~yyzhang/spring04
ECE 569 Database System Engineering
Spring 2004
Database Literature


Journals

IEEE Transaction on Knowledge and Data Engineering

ACM Transactions on Database Systems

VLDB Journal
Conferences

IEEE Data Engineering Conference

ACM SIGMDO

Very Large Database (VLDB)
ECE 569 Database System Engineering
Spring 2004
Time and Location


Core 538 schedule

Monday. 8:00 – 11:30, 1:40 – 4:30, after 5:50

Tuesday. Except 3:00 – 4:30

Wednesday. Before 4:30, after 6:00

Thursday. all day

Friday. Before 3, after 6.
Part time students please let me know (access to
Core, and 5th floor)
ECE 569 Database System Engineering
Spring 2004
Example – medical database

Entities in database, the types and names of their
attributes, and relationships between entities.
Date
Balance
Account
Made To
Billed
Patient
From
Address
ECE 569 Database System Engineering
Disease
Vital Sign
Time
Blood
Diagnosed
Treatment
Amount
Pulse
Name
Room#
Payment
Name
Spring 2004
System Architecture
DDL
Statements
Interactive
Queries
Application
Program
DDL Compiler
Query
Compiler
Precompiler
Parametric
Users
Host Compiler
DML
Execute
Data
Dictionary
DML Compiler
Runtime DB
Processor
File Manager
Compiled
Transaction
Execute
Concurrency control
and recovery
DDL: Data Definition Language
 DML: Data Manipulation Language

Stored
DB
ECE 569 Database System Engineering
Spring 2004