OLAP
Download as PPTX, PDF14 likes26,531 views
OLAP provides multidimensional analysis of large datasets to help solve business problems. It uses a multidimensional data model to allow for drilling down and across different dimensions like students, exams, departments, and colleges. OLAP tools are classified as MOLAP, ROLAP, or HOLAP based on how they store and access multidimensional data. MOLAP uses a multidimensional database for fast performance while ROLAP accesses relational databases through metadata. HOLAP provides some analysis directly on relational data or through intermediate MOLAP storage. Web-enabled OLAP allows interactive querying over the internet.
OLAP
- 1. OLAP V. Saranya AP/CSE Sri Vidya college of Engineering & Technology, Virudhunagar
- 2. Need for OLAP • Business problems need query centric db. • Need multidimensional approach. Characteristics of above problems: Extract large number of records from large data set. Data summary. To solve these kind of problems we need OLAP
- 3. Introduction to OLAP • Continuous iterative process. • Operations are: – Drill down – Drill up – pivot
- 4. Multidimensional data model • How many students done the conducted by department in college. • Dimensions are: exams – Students – Exams – Department – College. Response time of multidimensional query depends upon the number of cells to be added on the fly. Number of dimension increases=no of cube cell increase
- 5. Data Cubes
- 6. • • • • • • • • • • • • OLAP Guidelines Multidimensional conceptual view Transparency Accessibility Consistent reporting performance Client/Server architecture Generic dimensionality Dynamic square matrix handling Multiuser support Unrestricted cross dimensional operation Institutive data manipulation Flexible reporting Unlimited dimension and aggregation levels
- 7. • • • • Comprehensive database management tools The ability to drill down to detail view Incremental database refresh SQL interface.
- 8. Classification of OLAP tools • Based on multidimensional db. • Allow the users to analyze the data using views. • Need MDDB. • Classifications: – MOLAP – ROLAP
- 9. M(Multidimensional)OLAP • Uses MDDBMS to organize and navigate data • Data Structure: Array • Segregate the OLAP thro API Pros: Excellent performance Response time. Cons: Series analysis iteration
- 10. Example Organization tool • Arbor software: ESSbase • Oracle: Express server • Pilot Software: Light Slip Server • Snipper: TM/I • Planning Science: Gentium • Kenan technology: Multiway
- 11. Challenges • Data structure to support multiple subject area of data. • Analyze which data can be navigated and analyzed. • When the navigation changes the data structure needs to be physically reorganized. • Need different skill set and tools for DBA to build, maintain database. • Need hybrid solution. Hybrid Solution: Integration of multidimensional data storage with RDBMS, Provide users with MDDS Data maintained in RDBMS.
- 12. MOLAP Architecture Database Server Load MOLAP Server Info Request SQL Result Front End Tool Meta Data Request Processing Result Set
- 13. • This allows the MDDS to dynamically obtain the detail maintained in RDBMS when the application reaches the bottom of multidimensional cells during drill down analysis. • Best for Sensitive applications.
- 14. ROLAP • Fastest growing style of OLAP • Products of ROLAP have been engineered to support products directly through meta data. • Enables multi dimensional views of 2D relational tables. • Pros: – Flexibility • Cons: – Data base design
- 15. ROLAP ROLAP Server Database Server Info Request SQL Result set Front End Tool Meta Data Request Processing Result Set
- 16. • • • • Vendors Information advantage Microstrategy Platinum/Prodea software Sybase Tools Axsys Dss agent/ Dss server Beacon High gate project
- 17. Managed Query Environment/HOLAP • Provides user with ability to perform limited analysis capability either directly with RDBMS products or • Intermediate MOLAP. • The ad hoc query converted to provide data cube. Done by: 1. Convert the query to select data from DBMS 2. Deliver the data to desktop where it is placed in data cube. 3. Data cube is stored locally to reduce overhead of creation of the cube. 4. Now user can perform multi dimensional analysis.
- 18. HOLAP/MQE/Hybrid architecture SQL Query Database Server Result set OR Load RDB MS SQL Result set Info Request MOLAP Server Result Set Front End Tool
- 19. • Pros: – Simple installation – Administration is easy – Network traffic is less • Cons: – Redundancy – Inconsistency.
- 20. OLAP tools and Internet • Internet free resource, provides connectivity, can do complex administration jobs, store and manage data applications • Data warehousing General features of web enabled data access: • 1st generation websites: – Static distribution model – Client access static html pages via browser. – Decision support reports stored as html doc and delivered to users. • Deficiencies: – Interaction with clients.
- 21. • 2nd generation: – Supports interaction – Multi tiered architecture – Client submits the query in html to web server – Server transform the request to CGI – The gateway submits SQL queries to db and receives and translates to html and sends to page requester.