Transcript madgavkar
Tanvi Madgavkar CSE 7330 FALL 2009 Ralph Kimball states that : A data warehouse is a copy of transaction data specifically structured for query and analysis. Bill Inmon states that : A warehouse is a subject-oriented, integrated, time-variant and non-volatile collection of data in support of management's decision making process. • A data warehouse provides a common data model for all data of interest regardless of the data's source. • Prior to loading data into the data warehouse, inconsistencies are identified and resolved. • The information in the warehouse can be stored safely for extended periods of time. • It is a short for On Line Transaction Processing. • OLTP refers to a class of systems that facilitate and manage transaction-oriented applications, typically for data entry and information retrieval. • It is characterized by a large number of short on-line transactions. • The main emphasis for OLTP systems is put on very fast query environments. processing in multi-access • It is a short for On Line Analytical Processing. • OLAP is an approach to quickly multi-dimensional analytical queries. answer • The term OLAP was created as a slight modification of the traditional database term OLTP. • It is characterized by relatively low volume of transactions. In general, OLTP systems provide source data to data warehouses, whereas OLAP systems help to analyze it. OLTP OLAP Source of data OLTPs are the original source of data Data comes from various OLTP databases Purpose of data To run fundamental transaction related tasks To help with planning and decision support Queries Standardized and simple queries Complex queries involving Aggregation Processing Speed Very Fast Depends on the amount of data involved Space Requirements Relatively small Larger due to existence of historical data • Multidimensional OLAP - MOLAP This is the more traditional way of OLAP analysis. In MOLAP, data is not stored in the relational database but in a multidimensional cube. • Relational OLAP - ROLAP It works directly with relational databases, the base data is stored as relational tables and new tables are created to hold the aggregated information. • Hybrid OLAP - HOLAP HOLAP attempt to combine the advantages of MOLAP and ROLAP. Here, a database will divide data between relational to hold the larger quantities of detailed data and specialized storage for smaller quantities of less-detailed data. Steps in OLAP creation process: •OLAPs are designed to give an overview analysis of what happened. Hence the data storage has to be set up differently. •OLAP cubes also called a multidimensional cube or a hypercube data models. and are created from •OLAP cubes are not strictly cuboids - it is the name given to the process of linking data from the different dimensions. •There can be number of cubes, developed along units of dimensions or a giant cube can be formed with all the dimensions. •The OLAP cube is present at the core of any OLAP system and consists of number of tables arranged in a particular schema. • The cube metadata is typically created from either a star schema or snowflake schema of tables in a relational database. • The most common method is called the star design and it is called so, because it resembles a ‘star’ in shape. • The star schema also known as star join schema is the simplest style of data warehouse schema. • The star schema consists of a few fact tables, normally possibly only one, justifying the name referencing number of dimension tables. •Create Table FACT1 (time_key INTEGER, item_key INTEGER, branch_key INTEGER, Location_key INTEGER, PRIMARY KEY (time_key)) •Create Table TIME (time_key INTEGER, day VARCHAR(10), month VARCHAR(10), year VARCHAR(10), day_of_work VARCHAR(10), quarter VARCHAR(10), FOREIGN KEY time_key REFERENCES FACT1) •Create Table BRANCH (time_key INTEGER, branch_key INTEGER, branch_name VARCHAR(10), branch_type VARCHAR(10), FOREIGN KEY time_key REFERENCES FACT1) •Advantages: Simplest DW schema. Easy to understand. Easy to Navigate between the tables due to less number of joins. Most suitable for Query processing. •Disadvantages: Occupies more space. Highly Denormalized. • A snowflake schema is a logical arrangement of tables in a multidimensional database such that the entity relationship diagram resembles a snowflake in shape. • It is closely related to star schema as it is just a variation of it. The only difference being that dimensions are normalized into multiple related tables in a snowflake schema whereas the star schema's dimensions are denormalized with each dimension being represented by a single table. • Create Table FACT1 (time_key INTEGER, item_key INTEGER, branch_key INTEGER, Location_key INTEGER, PRIMARY KEY (time_key))) • Create Table ITEM(time_key INTEGER, item_key INTEGER, item_name VARCHAR(10), brand VARCHAR(10), type VARCHAR(10) , supplier_type VARCHAR(10) FOREIGN KEY time_key REFERENCES FACT1) • Create table SUPPLIER (time_key integer, supplier_key integer, supplier_type integer) FOREIGN KEY time_key REFERENCES FACT1) • Create Table FACT1 (time_key INTEGER, item_key INTEGER, branch_key INTEGER, Location_key INTEGER, PRIMARY KEY (time_key))) • Create Table LOCATION(time_key INTEGER, location_key INTEGER, street VARCHAR (10), city VARCHAR(10), PRIMARY KEY(location_key) FOREIGN KEY time_key REFERENCES FACT1) • Create table CITY (location_key INTEGER, city_key INTEGER, country VARCHAR (10), city VARCHAR (10), state VARCHAR (10))s FOREIGN KEY location_key REFERENCES LOCATION) • Advantages: These tables are easier to maintain. Saves the storage space. • Disadvantages: Due to large number of joins it is complex to navigate. • Star schema is a better option to choose from users point of view. This schema exposes users to the underlying table structures and also the queries are simpler in nature. It is more likely to be used when the data warehouse is large. •Snowflake schema are often better with more sophisticated query tools and smaller data warehouse. Even though its maintenance is relatively easy, it is based on environments having numerous queries with complex criteria and hence more query execution time. W.H. Inmon. “What is a Data Warehouse?, Prism, Volume 1, Number 1, 1995”. Ralph Kimball. “The Data Warehouse Toolkit: Practical Techniques for Building Dimensional Data Warehouses”. Jun Yang. “WareHouse Information Prototype at Stanford”. C. Caldeira. "Data Warehousing – Concepts and Models". RainMaker DataWarehousing. “OLAP_vs_OLTP.pdf”, http://www.rainmaker works.com/pdfdocs/OLTP_vs_OLAP.pdf “Data Warehousing: A look at Business Intelligence and Data Warehouse”, http://www.1keydata.com/dataware housing/ molap-rolap.html Hari Mailvaganam. “ Data Warehousing Review – Introduction to OLAP”, http://www.dwreview.com/OLAP /Introduction_OLAP.html Mri Sonam. “What is the difference between star schema and snow flake schema?”, http://www.geekinterview.com/question_details/38599 Wikipedia, The Free Encyclopedia. “Data Warehouse, OLAP, OLTP, Star Schema, Snowflake Schema”, http://en.wikipedia.org/wiki/Main_Page