Database Management Systems – Week 1 Comprehensive Notes
Introduction to Data vs. Information
Data
Raw, unprocessed facts, figures, observations, statistics and values.
Examples: individual exam scores, customer phone numbers, sensor readings.
Information
Data that has been processed or interpreted so it gains meaning and context for decision-making.
Key takeaway: Data ≠ Information; processing is required to convert the former into the latter.
Database & DBMS Fundamentals
Database
Structured record-keeping system that stores inter-related data sets.
Enables storage, retrieval and management of large volumes of information (e.g., Student, Subject, Faculty, Marks tables).
DBMS (Database Management System)
Software suite that creates, accesses, manages and secures databases.
Acts as the interface between application programs and physical data.
Core operations: Add/Insert, Modify/Update, Delete, Search, Report generation.
Why use a DBMS instead of simple files?
Minimises redundancy; enforces consistency; supports multiple users; provides security, backup & recovery; offers query language; optimises performance.
Core Characteristics of Modern DBMS
Treats data as real-world entities with attributes & behaviour.
Represents entities/relations as TABLES (relation-based model).
Data & application isolation via metadata; changes to data structure rarely require application rewrites.
Supports transactions (ACID), security, concurrency, query languages, and data independence.
Reduces redundancy, assures consistency, allows concurrent multi-user access.
Where Databases Appear (Real-World Domains)
Banking (accounts, transactions).
Airlines (reservations, schedules).
Universities (registrations, grades).
E-commerce & online retail (orders, recommendations).
Manufacturing & supply-chain (inventory, production, orders).
HR (employee records, payroll, tax).
Common DBMS Products
Commercial RDBMS: Oracle, Microsoft SQL Server, IBM DB2.
Open-source RDBMS: MySQL, PostgreSQL, MariaDB.
NoSQL / Document: MongoDB.
Desktop: Microsoft Access.
Types of Database Systems
RDBMS (Relational DBMS)
Data organised as tables (relations) with rows (tuples) and columns (attributes).
Uses primary keys (PK) & foreign keys (FK) to enforce relationships.
Example mini-schema:
Students(ID#, Name, Phone, DOB)
Courses(ClassID, Title, Credits)
Takes_Course(ID#, ClassID, Semester)
OODBMS (Object-Oriented DBMS)
Stores objects that encapsulate state (data) + behaviour (methods).
Can persist complex types: photos, signatures, audio, video, geometries.
ORDBMS (Object-Relational DBMS)
Hybrid; extends relational engine with object features (user-defined types, inheritance, multimedia columns).
Example domain: Bank with shared CITY/STATE/COUNTRY domain across Customer, Employee, Account tables.
Traditional File System Overview
Electronic data kept as independent files.
Flat file = single table in one file; searching requires row-by-row parsing into an in-memory array → inefficient.
File types in legacy processing:
Master files: rarely updated (e.g., Employee-Master).
Transaction files: day-to-day activity (e.g., Loan release records).
Table files: periodically updated reference tables (e.g., Interest Rate table).
Security/Recovery files: backup snapshots.
File System vs. Database System (Comparative)
Data redundancy & inconsistency: HIGH in FMS, MINIMAL in DBMS.
Data isolation/scattering: Present in FMS, eliminated in DBMS via unified schema.
Transactions & concurrency: Limited/none in FMS, full ACID & locking in DBMS.
Security: Basic OS-level in FMS vs granular privileges in DBMS.
Recovery: Manual & error-prone in FMS vs automated logging & backup in DBMS.
Problems Addressed by DBMS
Data Redundancy
Duplicate data across departments (e.g., student phone stored in Admin, Accounts, Exams) wastes space.
Data Inconsistency
Updates in one location not propagated to others → conflicting versions.
Concurrent Access & Recovery
Multiple users (ATMs) need simultaneous balance updates; DBMS locks/transactions preserve integrity.
Recovery subsystem tracks free disk blocks, directory trees, journals for crash restore.
Data Models
Hierarchical Model
Parent-child tree; single parent per child.
Fast, predictable access but rigid links; poor for many-to-many.
Example: College → Department → Programme → Student.
Network Model
Extends hierarchy: record can have multiple parents; supports 1:1, 1:N, M:N.
More flexible but link maintenance complex.
Example: ITEM related to both STORE and TRANSACTION segments.
Relational Model
Data stored in 2-D tables; SQL used for manipulation.
Advantages: structural flexibility, multiple user views, strong mathematical foundation.
Limitation: Distributed partitioning needs extra mechanisms.
Object-Oriented Model
Everything is an OBJECT with attributes & methods.
Allows user-defined access, custom data types; suited for complex multimedia data.
Challenges: Lacks universal standard model; fewer mature tools.
DBMS Architectures
One-Tier (Single-Tier)
Presentation + Business + Data in one machine (e.g., MS-Access, desktop games).
Two-Tier (Client/Server)
Client = GUI (presentation).
Server = combined Business logic + Database.
Clients communicate via API/ODBC for direct, faster calls.
Three-Tier (n-Tier)
Separate layers:
Presentation (browser/mobile).
Business/Application server (validation, calculations).
Database server (data persistence).
Used by web apps (e.g., Facebook); promotes scalability, multiple views, security.
Schema Levels & Data Abstraction
Internal (Physical): disk blocks, indices, file structures.
Conceptual (Logical): global schema; entities, relationships, constraints.
External (View): user-specific subsets or perspectives.
Benefit: Data Independence – ability to change one level without affecting others.
Core Components Inside a DBMS
Query Processor
DDL Interpreter: parses schema definitions → data dictionary.
DML Compiler: converts SQL to low-level evaluation plan.
Query Optimizer: chooses least-cost plan.
Query Evaluation Engine: executes plan.
Storage Manager
Interface between logical requests & physical data.
Sub-modules:
Authorization & Integrity Manager (security, constraints).
Transaction Manager (ACID, concurrency control).
File Manager (disk space allocation, directory structure).
Buffer Manager (caches disk blocks in RAM; vital for large datasets).
Disk Storage Elements
Data files: actual table rows.
Data dictionary: schema metadata.
Indices: auxiliary structures for fast lookup.
People in the DB Environment
Naïve Users: invoke predefined apps (ATM teller, web customer).
Application Programmers: write code; may use RAD 4GL tools.
Sophisticated Users: analysts using ad-hoc SQL/OLAP.
Specialized Users: build non-traditional apps (AI knowledge bases, graphics, audio DBs).
System Administrators: manage hardware/OS.
Database Designers/Analysts: model schema & constraints.
Database Administrators (DBA)
Select HW/DB products, set standards, tune performance, ensure security, backup/recovery, permissions.
Contemporary Challenges in DBMS Construction
Data Security
>100k publicly exposed DB instances hacked recently.
Must implement built-in DB security + organisational controls.
Performance
Growing data volumes with stricter response times; DB should deliver high native performance even on modest hardware.
Data Safety / Integrity
Must guarantee no data loss while maintaining throughput; balance between ACID and speed.
Resource Utilization
Optimise CPU, RAM, I/O to extract maximum work per nanosecond.
High Availability
Clustering, replication to avoid single-server downtime; akin to adding extra cash registers during rush hour.
Ethical & Practical Implications
Protecting confidential data (GDPR, HIPAA compliance) is both a legal and moral requirement.
Accurate, consistent data underpins fair decisions (e.g., loan approvals).
Efficient resource usage lowers energy footprint, contributing to sustainability.
Quick Numerical / Formal References
ACID properties:
Atomicity
Consistency
Isolation
Durability
Basic SQL selection: \text{SELECT } * \text{ FROM Students WHERE Age} > 18
Summary Checklist (Week 1)
Differentiated Data vs Information.
Defined Database, DBMS, RDBMS, OODBMS, ORDBMS.
Analysed File types and limitations vs DBMS strengths.
Explored four classical Data Models.
Detailed 1-, 2- and 3-tier Architectures; Internal–Conceptual–External schemas.
Mapped DBMS internals: Query Processor, Storage Manager, Disk structures.
Identified user roles & DBA responsibilities.
Highlighted five modern DB challenges: Security, Performance, Safety, Resource, Availability.
Use these structured notes as a comprehensive replacement for the original Week-1 slides and lectures.