1. OLAP – On Line Analytical Processing

2. Session Objectives Objectives:
At the end of this session, you will be able to:
Define On Line Analytical Processing
Understand the need for OLAP and applications of OLAP in BI
Describe the various OLAP solutions and Architecture
Comparison of different OLAP architectures
Evaluation parameters to be considered for selecting an OLAP tool

3. What is OLAP?
OLAP (On Line Analytical Processing) applications - designed for online ad-hoc data access and analysis.
Data organized into multiple dimensions.
Access to analytical content such as time series and trend analysis views and summary level information.
A set of functionality that attempts to facilitate multidimensional analysis.
Offers drill-down, drill-across and slice and dice capabilities.

4. OLAP - Fast Analysis On Line No piles of paper, please!
Analytical Establish patterns
Processing Data-based
Fast Analysis of Shared Multidimensional Information
On-line Analytical Processing (OLAP) is a category of software technology that enables analysts, managers and executives to gain insight into data through fast, consistent, interactive access to a wide variety of possible views of information.
OLAP is implemented in a multi-user client/server mode and offers consistently rapid response to queries, regardless of database size and complexity.
OLAP helps the user synthesize enterprise information through comparative, personalized viewing, as well as through analysis of historical and projected data in various "what-if" data model scenarios. This is achieved through use of an OLAP Server.

5. Need for OLAP Dimensions can we think in ?
E.g. analysis by branch, product, agent, year !!!
2 or 3
Types of values we can handle ?
E.g. Sales, Profit, Cost
1 or 2
How many levels can we handle ?
E.g. number of products we can analyze
Spread-sheet view of 1 value type
user definable row
user definable column
Different graphical views
Cross-tab, pie chart,
Drilling down , level by level
Country, Region, Branch, Agents

6. Need for OLAP Many parameters affect a Measure (value)
e.g Sales influenced by product, region, time, distribution channel, etc.,
Linear analysis = reports
Many totals are at one level
Difficult to identify the key parameters

7. OLAP in an Enterprise OLAP Server
An OLAP server is a high capacity, multi-user data manipulation engine specifically designed to support and operate on multi-dimensional data structures. A multi- dimensional structure is arranged so that every data item is located and accessed based on the intersection of the dimension members, which define that item. The design of the server and the structure of the data are optimized for rapid ad-hoc information retrieval in any orientation, as well as for fast, flexible calculation and transformation of raw data based on formulaic relationships.
OLAP Client
End user applications that can request information from OLAP servers and provide two- dimensional or multi-dimensional displays, user modifications, selections, ranking, calculations, etc., for visualization and navigation purposes. OLAP clients may be as simple as a spreadsheet program retrieving a slice for further work by a spreadsheet- literate user or as high-functioned as a financial modeling or sales analysis application

8. Uses of OLAP Departments: Finance Marketing Sales Manufacturing
Analytical Capabilities:
Used by analysts and managers.
Offers aggregated view of the data, such as total revenues by customer profile, by product line, by geographical regions.

9. Functionality of OLAP Tools
Provides the decision support front-end for data warehousing.
Advanced statistical, financial, and analytical calculations.
Appropriate tools to access data from a relational database.
Appropriate tools to access or manage multidimensional data.
OLAP functionality is characterized by dynamic multi-dimensional analysis of consolidated enterprise data supporting end user analytical and navigational activities including:
Calculations and modeling applied across dimensions, through hierarchies and/or across members
Trend analysis over sequential time periods
Slicing subsets for on-screen viewing
Drill-down to deeper levels of consolidation
Reach-through to underlying detail data

10. Features of OLAP Applications
OLAP analytical features
Multi-dimensional views of data
Calculation intensive capabilities
Time intelligence
The OLAP Calculation engine in OLAP tools have a wide range of
built-in calculations such as:
Ratios
Time calculations
Statistics
Ranking
Custom formulas/algorithms
Forecasting and modeling
Multi dimensional views
A dimension represents one variable. A sales data might include at least these five dimensions:
Time, Product, Customer, Salesperson, Sales
A simple query using above dimensions can be – “ How much revenue did Ram generated in January from all the products he sold to Ram & Co. ?”
Calculation intensive
OLAP does more than simple aggregation. It can be used for share calculations (Percentage of total) and allocations(which use hierarchies from a top-down perspective), trend analysis and forecasting using historical data.
Time intelligence
Time is a unique dimension being sequential in character. Business performance is almost always judged over time, for e.g., this month Vs last month, this month Vs the same month last year. Concepts such as year-to-date and period over period comparisons are used in an OLAP

11. Evolution of OLAP

12. Star Schema
A Star Schema is a dimensional model created by mapping data entities from operational systems
It has a central table (fact table) that links all the other tables (dimension tables) together
Dimension: The same category of information. For example, year, month, day, and week are all part of the Time Dimension.
Measure: The property that can be summed or averaged using pre computed aggregates.

13. Facts and Measures Sales Revenue Gross Margin Net Profit Cost
Profitability
Facts or Measures are the Key Performance Indicators of an enterprise
Factual data about the subject area
Numeric, summarized

14. Dimension Sales Revenue What was sold ? (Measure)
Whom was it sold to ?
When was it sold ?
Where was it sold ?
Sales Revenue
(Measure)
Dimensions put measures in perspective
What, when and where qualifiers to the measures
Dimensions could be products, customers, time, geography etc.

15. Star Schema

16. Star Schema Example

17. Star Schema with Sample Data

18. CUBE
Cube
Multi dimensional databases store information in the form of cubes.
A cube is a collection of facts and related dimensions stored together in arrays.
Geography
Sales
A group of data cells arranged by the dimensions of the data. For example, a spreadsheet exemplifies a two-dimensional array with the data cells arranged in rows and columns, each being a dimension.
A three-dimensional array can be visualized as a cube with each dimension forming a side of the cube, including any slice parallel with that side.
Higher dimensional arrays have no physical metaphor, but they organize the data in the way users think of their enterprise. Typical enterprise dimensions are time, measures, products, geographical regions, sales channels, etc. HR
Time
Product

19. Basic Terminology of a Cube
Hierarchy: A hierarchy defines the navigating path for drilling up and drilling down. All attributes in a hierarchy belong to the same dimension.
Levels: These are organized into one or more hierarchies, typically from a coarse-grained level (for example, Year) down to the most detailed one (for example, Day).
Members: The individual category values (for example, 2002 or 21Jan2002).
Measures: These are the data values that are summarized and analyzed. Examples of measures are sales figures or operational costs.
Cells: These are the intersection of one member for every dimension and store the data for measures.

20. Basic Terminology of a Cube
Dimensions consist of
Dimension Name
Level
Hierarchy
Member
Time
Level Of
Detail
YEAR
QUARTER
1999
2000
2001 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2

21. Aggregates
Add up amounts for day 1
In SQL: SELECT sum(amt) FROM SALE
WHERE date = 1 81

22. Aggregates
Add up amounts by day
In SQL: SELECT date, sum(amt) FROM SALE
GROUP BY date

23. Another Example
Add up amounts by day, product
In SQL: SELECT date, sum(amt) FROM SALE
GROUP BY date, prodId
rollup
drill-down

24. Aggregates
Operators: sum, count, max, min, median and avg
“Having” clause
Using dimension hierarchy
average by region (within store)
maximum by month (within date)

25. Multi-dimensional cube:
The MOLAP Cube
Fact table view:
Multi-dimensional cube:
dimensions = 2

26. Multi-dimensional cube:
3-D Cube
Fact table view:
Multi-dimensional cube:
day 2
day 1
dimensions = 3

27. Example Dimensions: Time, Product, Store Attributes:
roll-up to region
Dimensions:
Time, Product, Store
Attributes:
Product (upc, price, …)
Store … …
Hierarchies:
Product  Brand  …
Day  Week  Quarter
Store  Region  Country NY
Store SF
roll-up to brand LA
Juice
Milk
Coke
Cream
Soap
Bread 10 34 56 32 12
Product
roll-up to week
M T W Th F S S
Time
56 units of bread sold in LA on M

28. Cube Aggregation: Roll-up
Example: computing sums
day 2
. . .
day 1
129
drill-down
rollup

29. Aggregation Using Hierarchies
day 2
store
day 1
region
country
(store s1 in Region A;
stores s2, s3 in Region B)

30. In SQL: SELECT * FROM SALE WHERE date = 1
Slicing
In SQL: SELECT * FROM SALE WHERE date = 1
day 2
day 1
TIME = day 1

31. OLAP Solutions and Architecture

32. OLAP - Classification
Online Analytical Processing (OLAP) can be done on:
Relational databases
Multidimensional databases
OLAP products are grouped into three categories:
Relational OLAP (ROLAP)
Multidimensional OLAP (MOLAP)
Hybrid OLAP (HOLAP)

33. MOLAP Multi-dimensional OLAP
Geography
Age Group
Brand
Multi-dimensional OLAP
MOLAP is a technology which uses a multi-dimensional database that stores data as n-dimensional cube
Each face of a multi-dimensional cube represents a business dimension. Every point of intersection of dimensions, called a cell, represents a business fact relating to the dimensions.Practically MOLAP can store more than three dimensions in a cube.

34. Architecture of MOLAP Cube Size Critical LAN Router Firewall Issues:
non-live connection
Used for updating the MOLAP data cube only
LAN
Desktop Systems
Data Mart Server
MOLAP Client
Tools
RDBMS
Connectivity Middleware
MOLAP Server
MDDBMS/Data Cube
MOLAP Application
Router
Firewall
Multidimensional data bases are also known as MDDB or MDDBS .A class of proprietary, non-relational database management tools that store and manage data in a multidimensional manner. MOLAP servers directly store multidimensional data in special data structures like arrays or cubes.
Source data is pre-consolidated. MDDs pre-calculate and store every measure at every hierarchy summary level at load time. MOLAP servers physically stage the processed multi-dimensional information to deliver consistent and rapid response times to end users
Multidimensional OLAP analysis usually requires proprietary (non-SQL) access tools. MOLAP often utilizes a 3-Tier environment, where middle tier server preprocesses data from an RDBMS. Some OLAP tools access an RDBMS directly and build cubes as a fat client.
Issues:
Size of Data Cube
Cubes deployment
Size of Update Data Set
Intranet
Internet
Thin Clients
WWW Browser

35. MOLAP Products Oracle's Oracle Express Server
Cognos - Powerplay Transformer
Essbase (Hyperion Software)
Holos (Seagate Software)
There are also lot of MOLAP vendors in the market.
The main features of MOLAP products are :
Forecasting / Budgeting
Summarized as well as detailed data for analysis
Highly complex calculations
Infrequent / Small Updates
Three / N tier architecture
Upsides of MOLAP are :
Instant response
Value add functions (ranking, % change)
Provides more intuitive commands for performing drill down and other analytic operations.
A single cube holds answers to a number of requests

36. Architecture of ROLAP Issues: Aggregate Awareness Response Time
LAN
Data Mart Server
ROLAP Server
Desktop Systems
RDBMS
Connectivity Middleware
ROLAP Application
ROLAP Client
Tools
Router /
Firewall
Issues:
Aggregate Awareness
Response Time
Network Capacity
In ROLAP data is stored in the form of records in relational databases. It also supports extensions to SQL and ad-hoc queries. ROLAP servers efficiently implement the multidimensional data model and operations in a relational database.
ROLAP Uses RDBMS tables as data source .Architecture can be designed as 3/N tier system. The Calculation engine may be located anywhere.RDBMSs supports star / Snowflake schema. Supports various data types such as numeric, textual, spatial, audio, graphic, and video data. Different types of indexing mechanisms are available in order to improve query performance.
Features of ROLAP :
ROLAP servers are highly scaleable, in both number of users and amount of data
Calculation-on-the-fly.
Can be run from the client, the data-base server or a middle tier
Limitless filtering and grouping
Can handle extremely large number of dimensions
Intranet
Internet
Thin Clients
WWW Browser

37. ROLAP Products Brio Query Enterprise Business Objects Metacube
DSS Server
Information Advantage
The specified tools are market leaders in ROLAP.
Generally ROLAP architecture offers these features :
Analysis of transaction detail data
Very Large Database
Frequent / Large Updates
Aggregate aware
Monitoring
Three / N tier architecture
Upsides of ROLAP are :
Leverage RDBMS capability
Value add functions (ranking, % change)
No additional loads
No additional data sets to manage
Scalability

38. Architecture of HOLAP Issues: Cube elements Integration with RDBMS LAN
MOLAP Server
ROLAP Server
Desktop Systems
MDDBMS/Data Cube
MOLAP Application
ROLAP Application
HOLAP Client
Tools
Router/Fire
wall
HOLAP (Hybrid OLAP) is a mix of MOLAP and relational architecture that supports queries against summary and transaction data in an integrated fashion.HOLAP uses a multi-dimensional database server as middle ware to access data stored in a relational database.
It provides users with the detailed transaction data that contributes to the summary totals stored in the multidimensional server.A relational database stores most of the data. A separate multi-dimensional database stores the most of the dense data, which is typically a small proportion of the total volume of data.
HOLAP architectures are typically more complex to implement and administer than ROLAP or MOLAP architectures
Issues:
Cube elements
Integration with
RDBMS

39. HOLAP Products Holos (Seagate Software)
Microsoft SQL Server OLAP Services
Pilot Software's Pilot Decision Support Suite
SAS

40. MOLAP Vs ROLAP

41. Comparison of Architectures

42. Strength and Weakness of MOLAP/ROLAP

43. Strength and Weakness of MOLAP/ROLAP
On-line Analytical Processing (OLAP) is a category of software technology that enables analysts, managers and executives to gain insight into data through fast, consistent, interactive access to a wide variety of possible views of information.
OLAP is implemented in a multi-user client/server mode and offers consistently rapid response to queries, regardless of database size and complexity.
OLAP helps the user synthesize enterprise information through comparative, personalized viewing, as well as through analysis of historical and projected data in various "what-if" data model scenarios. This is achieved through use of an OLAP Server.

44. Session Summary In this session, We have
Understood the need for OLAP and significance of Multidimensional analysis in a Data Warehouse.
Discussed about the evolution of OLAP.
Explained architectures, characteristics as well as the merits and
demerits of various OLAP solutions.

45. Thank you