Differences and Similarities: Data Warehouse vs Mulitdatabase - PowerPoint PPT Presentation
1 / 19
Title:
Differences and Similarities: Data Warehouse vs Mulitdatabase
Description:
Extractor/ Monitor. Extractor/ Monitor. Extractor/ Monitor. Architectur Similarities ... Extractor/ Monitor. Extractor/ Monitor. Extractor/ Monitor. Application ... – PowerPoint PPT presentation
Number of Views:685
Avg rating:3.0/5.0
Title: Differences and Similarities: Data Warehouse vs Mulitdatabase
1
Differences and Similarities Data Warehouse vs
Mulitdatabase
- Ove Granberg
- Thomas Sveen
2
Contents
- Introduction
- Architecture Differences Similarities
- Design Differences Similarities
3
Data Warehousing Architecture
Data Warehouse Server
Information Sources
OLAP Servers
Clients
4
Data Warehouse Architecture
- Characteristics
- OLAP ? Business analyst
- Index / join tech
- Multidimentional space
- Historical
Clients
5
MDBSs Architecture
- Query Driven
- Types
- Federated
- Un-federated
- Distinguises
- Homogenous / hetrogenous
- MDBSs ? Autonomy sites
Clients
6
Architectur
Clients
- Similarities
- Sources
- Relational db
- Heterogenous /homogenous
- Distributed
Data
- Differences
- Dw ? flatt files
Sources
Multidatabase
Data Warehouse
7
Architectur Similarities
Clients
- Centralized
- Autonomous sites
- C/S approach
- Local users
- (Local schemas)
- OLTP
Data
Multidatabase
Data Warehouse
8
Architectur Similarities
Clients
- External user support
Data
Multidatabase
Data Warehouse
9
Architectur Similarities
Clients
Data
- Multidatabase
Multidatabase
Data Warehouse
10
Design
- Data/Database Integration
- Data modeling
11
Design
- Data/Database Integration
- Goal
- Uniform data model (schema)
- To steps
- Schema Translation
- Schema conflict, Semantic conflict,
- Schema Integration
12
Data/Database Integration
- Differences?
- Data warehouse
- OLAP
- Queries are long and complex
- Historical data
- Read only
- Hundreds of users
- Predictable queries
- Multidatabase
- OLTP
- Many smal transactions
- Current data
- Read / Write
- Tousands of users
- Unpredictable queries
13
Data/Database Integration
- Data warehouse
- In advance, Eager
- Query
- Directly
- Multidatabase
- On demand, Lazy
- Query
- Through global schema
14
Data Warehouse
Queries
Clients
Load phase
Integration phase
Refresh
Extraction/Translation phase
15
MDBS
Queries
Clients
Integration phase
Translation phase
16
Data/Database Integration
- Advantages
- Data warehouse
- High query performance
- Do not interfere with local processing at the
sources - Historical data
- Restructure, Adjust data
- Multidatabase
- High transaction performance
- Unpredictable queries
- Data - Up to date
17
Data/Database Integration
- Multidatabase
- Delays
- Time consuming
- Competes with local users
- Poor query capability
- Disadvantages
- Data warehouse
- Data out of date
- Extra storage space
- Administrator must specify data that should be
extracted
18
Data Modeling
- Data warehouse
- Dimensonal model
- Star Schema
- Snowflake Schema
- Denormalized
- Summarized, Adjusted data
- Matrialized view
- Multidatabase
- ER model
- Normalized
- Raw data
19
Summary
- Differences and Similarities
- Data warehouse vs Multidatabase
- Database Integration
- Datamodeling
