ETL and OLAP Systems

Transcription

1 ETL and OLAP Systems Krzysztof Dembczyński Intelligent Decision Support Systems Laboratory (IDSS) Poznań University of Technology, Poland Software Development Technologies Master studies, first semester Academic year 2017/18 (winter course) 1 / 40

2 Review of the Previous Lecture Mining of massive datasets. Evolution of database systems. Dimensional modeling: Three goals of the logical design of data warehouse: simplicity, expressiveness and performance. The most popular conceptual schema: star schema. Designing data warehouses is not an easy task... 2 / 40

3 Outline 1 Motivation 2 ETL 3 OLAP Systems 4 Summary 3 / 40

4 Outline 1 Motivation 2 ETL 3 OLAP Systems 4 Summary 4 / 40

5 Motivation OLAP queries are usually performed in a separate system, i.e., a data warehouse. 5 / 40

6 Motivation OLAP queries are usually performed in a separate system, i.e., a data warehouse. Transferring data to data warehouse: 5 / 40

7 Motivation OLAP queries are usually performed in a separate system, i.e., a data warehouse. Transferring data to data warehouse: Data warehouses combine data from multiple sources. 5 / 40

8 Motivation OLAP queries are usually performed in a separate system, i.e., a data warehouse. Transferring data to data warehouse: Data warehouses combine data from multiple sources. Data must be translated into a consistent format. 5 / 40

9 Motivation OLAP queries are usually performed in a separate system, i.e., a data warehouse. Transferring data to data warehouse: Data warehouses combine data from multiple sources. Data must be translated into a consistent format. Data integration represents 80% of effort for a typical data warehouse project! 5 / 40

10 Motivation OLAP queries are usually performed in a separate system, i.e., a data warehouse. Transferring data to data warehouse: Data warehouses combine data from multiple sources. Data must be translated into a consistent format. Data integration represents 80% of effort for a typical data warehouse project! Optimization of data warehouse: 5 / 40

11 Motivation OLAP queries are usually performed in a separate system, i.e., a data warehouse. Transferring data to data warehouse: Data warehouses combine data from multiple sources. Data must be translated into a consistent format. Data integration represents 80% of effort for a typical data warehouse project! Optimization of data warehouse: Data storage: relational or multi-dimensional. 5 / 40

12 Motivation OLAP queries are usually performed in a separate system, i.e., a data warehouse. Transferring data to data warehouse: Data warehouses combine data from multiple sources. Data must be translated into a consistent format. Data integration represents 80% of effort for a typical data warehouse project! Optimization of data warehouse: Data storage: relational or multi-dimensional. Additional data structures: sorting, indexing, summarizing, cubes. 5 / 40

13 Motivation OLAP queries are usually performed in a separate system, i.e., a data warehouse. Transferring data to data warehouse: Data warehouses combine data from multiple sources. Data must be translated into a consistent format. Data integration represents 80% of effort for a typical data warehouse project! Optimization of data warehouse: Data storage: relational or multi-dimensional. Additional data structures: sorting, indexing, summarizing, cubes. Refreshing of data structures. 5 / 40

14 Motivation OLAP queries are usually performed in a separate system, i.e., a data warehouse. Transferring data to data warehouse: Data warehouses combine data from multiple sources. Data must be translated into a consistent format. Data integration represents 80% of effort for a typical data warehouse project! Optimization of data warehouse: Data storage: relational or multi-dimensional. Additional data structures: sorting, indexing, summarizing, cubes. Refreshing of data structures. Querying multidimensional data: 5 / 40

15 Motivation OLAP queries are usually performed in a separate system, i.e., a data warehouse. Transferring data to data warehouse: Data warehouses combine data from multiple sources. Data must be translated into a consistent format. Data integration represents 80% of effort for a typical data warehouse project! Optimization of data warehouse: Data storage: relational or multi-dimensional. Additional data structures: sorting, indexing, summarizing, cubes. Refreshing of data structures. Querying multidimensional data: SQL extensions, 5 / 40

16 Motivation OLAP queries are usually performed in a separate system, i.e., a data warehouse. Transferring data to data warehouse: Data warehouses combine data from multiple sources. Data must be translated into a consistent format. Data integration represents 80% of effort for a typical data warehouse project! Optimization of data warehouse: Data storage: relational or multi-dimensional. Additional data structures: sorting, indexing, summarizing, cubes. Refreshing of data structures. Querying multidimensional data: SQL extensions, Multidimensional expressions (MDX), 5 / 40

17 Motivation OLAP queries are usually performed in a separate system, i.e., a data warehouse. Transferring data to data warehouse: Data warehouses combine data from multiple sources. Data must be translated into a consistent format. Data integration represents 80% of effort for a typical data warehouse project! Optimization of data warehouse: Data storage: relational or multi-dimensional. Additional data structures: sorting, indexing, summarizing, cubes. Refreshing of data structures. Querying multidimensional data: SQL extensions, Multidimensional expressions (MDX), Map-reduce-based language. 5 / 40

18 Outline 1 Motivation 2 ETL 3 OLAP Systems 4 Summary 6 / 40

19 ETL ETL = Extraction, Transformation, and Load Extraction of data from source systems, Transformation and integration of data into a useful format for analysis, Load of data into the warehouse and build of additional structures. Refreshment of data warehouse is closely related to ETL process. The ETL process is described by metadata stored in data warehouse. Architecture of data warehousing: Data sources Data staging area Data warehouse 7 / 40

20 ETL 8 / 40

21 Tools for ETL Data extraction from heterogeneous data sources. Data transformation, integration, and cleansing. Data quality analysis and control. Data loading. High-speed data transfer. Data refreshment. Managing and analyzing metadata. Examples of ETL tools: MS SQL Server Integration Services(SSIS), IBM Infosphere DataStage, SAS ETL Studio, Oracle Warehouse Builder, Oracle Data Integrator, Business Objects Data Integrator, Pentaho Data Integration. 9 / 40

22 Tools for ETL MS SQL Server Integration Services(SSIS) 10 / 40

23 Tools for ETL MS SQL Server Integration Services(SSIS) 11 / 40

24 Data extraction Data warehouse needs extraction of data from different external data sources: 12 / 40

25 Data extraction Data warehouse needs extraction of data from different external data sources: operational databases (relational, hierarchical, network, itp.), 12 / 40

26 Data extraction Data warehouse needs extraction of data from different external data sources: operational databases (relational, hierarchical, network, itp.), files of standard applications (Excel, COBOL applications), 12 / 40

27 Data extraction Data warehouse needs extraction of data from different external data sources: operational databases (relational, hierarchical, network, itp.), files of standard applications (Excel, COBOL applications), additional databases (direct marketing databases) and data services (stock data), 12 / 40

28 Data extraction Data warehouse needs extraction of data from different external data sources: operational databases (relational, hierarchical, network, itp.), files of standard applications (Excel, COBOL applications), additional databases (direct marketing databases) and data services (stock data), various log files, 12 / 40

29 Data extraction Data warehouse needs extraction of data from different external data sources: operational databases (relational, hierarchical, network, itp.), files of standard applications (Excel, COBOL applications), additional databases (direct marketing databases) and data services (stock data), various log files, and other documents (.txt,.doc, XML, WWW). 12 / 40

30 Data extraction Data warehouse needs extraction of data from different external data sources: operational databases (relational, hierarchical, network, itp.), files of standard applications (Excel, COBOL applications), additional databases (direct marketing databases) and data services (stock data), various log files, and other documents (.txt,.doc, XML, WWW). Access to data sources can be difficult: 12 / 40

31 Data extraction Data warehouse needs extraction of data from different external data sources: operational databases (relational, hierarchical, network, itp.), files of standard applications (Excel, COBOL applications), additional databases (direct marketing databases) and data services (stock data), various log files, and other documents (.txt,.doc, XML, WWW). Access to data sources can be difficult: Data sources are often operational systems, providing the lowest level of data. 12 / 40

32 Data extraction Data warehouse needs extraction of data from different external data sources: operational databases (relational, hierarchical, network, itp.), files of standard applications (Excel, COBOL applications), additional databases (direct marketing databases) and data services (stock data), various log files, and other documents (.txt,.doc, XML, WWW). Access to data sources can be difficult: Data sources are often operational systems, providing the lowest level of data. Data sources are designed for operational use, not for decision support, and the data reflect this fact. 12 / 40

33 Data extraction Data warehouse needs extraction of data from different external data sources: operational databases (relational, hierarchical, network, itp.), files of standard applications (Excel, COBOL applications), additional databases (direct marketing databases) and data services (stock data), various log files, and other documents (.txt,.doc, XML, WWW). Access to data sources can be difficult: Data sources are often operational systems, providing the lowest level of data. Data sources are designed for operational use, not for decision support, and the data reflect this fact. Multiple data sources are often from different systems, run on a wide range of hardware and much of the software is built in-house or highly customized. 12 / 40

34 Data extraction Data warehouse needs extraction of data from different external data sources: operational databases (relational, hierarchical, network, itp.), files of standard applications (Excel, COBOL applications), additional databases (direct marketing databases) and data services (stock data), various log files, and other documents (.txt,.doc, XML, WWW). Access to data sources can be difficult: Data sources are often operational systems, providing the lowest level of data. Data sources are designed for operational use, not for decision support, and the data reflect this fact. Multiple data sources are often from different systems, run on a wide range of hardware and much of the software is built in-house or highly customized. Data sources can be designed using different logical structures. 12 / 40

35 Data extraction Identification of concepts and objects does not have to be easy. 13 / 40

36 Data extraction Identification of concepts and objects does not have to be easy. Example: Extract information about sales from the source system. 13 / 40

37 Data extraction Identification of concepts and objects does not have to be easy. Example: Extract information about sales from the source system. What is meant by the term sale? A sale has occurred when 13 / 40

38 Data extraction Identification of concepts and objects does not have to be easy. Example: Extract information about sales from the source system. What is meant by the term sale? A sale has occurred when 1 the order has been received by a customer, 13 / 40

39 Data extraction Identification of concepts and objects does not have to be easy. Example: Extract information about sales from the source system. What is meant by the term sale? A sale has occurred when 1 the order has been received by a customer, 2 the order is sent to the customer, 13 / 40

40 Data extraction Identification of concepts and objects does not have to be easy. Example: Extract information about sales from the source system. What is meant by the term sale? A sale has occurred when 1 the order has been received by a customer, 2 the order is sent to the customer, 3 the invoice has been raised against the order. 13 / 40

41 Data extraction Identification of concepts and objects does not have to be easy. Example: Extract information about sales from the source system. What is meant by the term sale? A sale has occurred when 1 the order has been received by a customer, 2 the order is sent to the customer, 3 the invoice has been raised against the order. It is a common problem that there is no table SALES in the operational databases; some other tables can exist like ORDER with an attribute ORDER STATUS. 13 / 40

42 Conflicts and dirty data Different logical models of operational sources, 14 / 40

43 Conflicts and dirty data Different logical models of operational sources, Different data types (account number stored as String or Numeric), 14 / 40

44 Conflicts and dirty data Different logical models of operational sources, Different data types (account number stored as String or Numeric), Different data domains (gender: M, F, male, female, 1, 0), 14 / 40

45 Conflicts and dirty data Different logical models of operational sources, Different data types (account number stored as String or Numeric), Different data domains (gender: M, F, male, female, 1, 0), Different date formats (dd-mm-yyyy or mm-dd-yyyy), 14 / 40

46 Conflicts and dirty data Different logical models of operational sources, Different data types (account number stored as String or Numeric), Different data domains (gender: M, F, male, female, 1, 0), Different date formats (dd-mm-yyyy or mm-dd-yyyy), Different field lengths (address stored by using 20 or 50 chars), 14 / 40

47 Conflicts and dirty data Different logical models of operational sources, Different data types (account number stored as String or Numeric), Different data domains (gender: M, F, male, female, 1, 0), Different date formats (dd-mm-yyyy or mm-dd-yyyy), Different field lengths (address stored by using 20 or 50 chars), Different naming conventions: homonyms and synonyms, 14 / 40

48 Conflicts and dirty data Different logical models of operational sources, Different data types (account number stored as String or Numeric), Different data domains (gender: M, F, male, female, 1, 0), Different date formats (dd-mm-yyyy or mm-dd-yyyy), Different field lengths (address stored by using 20 or 50 chars), Different naming conventions: homonyms and synonyms, Inconsistent information concerning the same object, 14 / 40

49 Conflicts and dirty data Different logical models of operational sources, Different data types (account number stored as String or Numeric), Different data domains (gender: M, F, male, female, 1, 0), Different date formats (dd-mm-yyyy or mm-dd-yyyy), Different field lengths (address stored by using 20 or 50 chars), Different naming conventions: homonyms and synonyms, Inconsistent information concerning the same object, Information concerning the same object, but indicated by different keys, 14 / 40

50 Conflicts and dirty data Different logical models of operational sources, Different data types (account number stored as String or Numeric), Different data domains (gender: M, F, male, female, 1, 0), Different date formats (dd-mm-yyyy or mm-dd-yyyy), Different field lengths (address stored by using 20 or 50 chars), Different naming conventions: homonyms and synonyms, Inconsistent information concerning the same object, Information concerning the same object, but indicated by different keys, Missing values, 14 / 40

51 Conflicts and dirty data Different logical models of operational sources, Different data types (account number stored as String or Numeric), Different data domains (gender: M, F, male, female, 1, 0), Different date formats (dd-mm-yyyy or mm-dd-yyyy), Different field lengths (address stored by using 20 or 50 chars), Different naming conventions: homonyms and synonyms, Inconsistent information concerning the same object, Information concerning the same object, but indicated by different keys, Missing values, / 40

52 Deduplication and householding Deduplication ensures that one accurate record exists for each business entity represented in a database, 15 / 40

53 Deduplication and householding Deduplication ensures that one accurate record exists for each business entity represented in a database, Householding is the technique of grouping individual customers by the household or organization of which they are a member; this technique has some interesting marketing implications, and can also support cost-saving measures of direct advertising. 15 / 40

54 Deduplication and householding Deduplication ensures that one accurate record exists for each business entity represented in a database, Householding is the technique of grouping individual customers by the household or organization of which they are a member; this technique has some interesting marketing implications, and can also support cost-saving measures of direct advertising. Example: 15 / 40

55 Deduplication and householding Deduplication ensures that one accurate record exists for each business entity represented in a database, Householding is the technique of grouping individual customers by the household or organization of which they are a member; this technique has some interesting marketing implications, and can also support cost-saving measures of direct advertising. Example: Consider the following rows in a database: Tim Jones 123 Main Street Marlboro MA T. Jones 123 Main St. Marlborogh MA Timothy Jones 321 Maine Street Marlborog AM Jones, Timothy 123 Maine Ave Marlborough MA / 40

56 Deduplication and householding Deduplication ensures that one accurate record exists for each business entity represented in a database, Householding is the technique of grouping individual customers by the household or organization of which they are a member; this technique has some interesting marketing implications, and can also support cost-saving measures of direct advertising. Example: Consider the following rows in a database: Tim Jones 123 Main Street Marlboro MA T. Jones 123 Main St. Marlborogh MA Timothy Jones 321 Maine Street Marlborog AM Jones, Timothy 123 Maine Ave Marlborough MA The sales for around $500 are counted for each tuple. 15 / 40

57 Deduplication and householding Deduplication ensures that one accurate record exists for each business entity represented in a database, Householding is the technique of grouping individual customers by the household or organization of which they are a member; this technique has some interesting marketing implications, and can also support cost-saving measures of direct advertising. Example: Consider the following rows in a database: Tim Jones 123 Main Street Marlboro MA T. Jones 123 Main St. Marlborogh MA Timothy Jones 321 Maine Street Marlborog AM Jones, Timothy 123 Maine Ave Marlborough MA The sales for around $500 are counted for each tuple. Is it the same person? 15 / 40

58 Load of data After extracting, cleaning and transforming, data must be loaded into the warehouse. 16 / 40

59 Load of data After extracting, cleaning and transforming, data must be loaded into the warehouse. Loading the warehouse includes some other processing tasks: checking integrity constraints, sorting, summarizing, creating indexes, etc. 16 / 40

60 Load of data After extracting, cleaning and transforming, data must be loaded into the warehouse. Loading the warehouse includes some other processing tasks: checking integrity constraints, sorting, summarizing, creating indexes, etc. Batch (bulk) load utilities are used for loading. 16 / 40

61 Load of data After extracting, cleaning and transforming, data must be loaded into the warehouse. Loading the warehouse includes some other processing tasks: checking integrity constraints, sorting, summarizing, creating indexes, etc. Batch (bulk) load utilities are used for loading. A load utility must allow the administrator to monitor status, to cancel, suspend, and resume a load, and to restart after failure with no loss of data integrity. 16 / 40

62 Data warehouse refreshment Refreshing a warehouse means propagating updates on source data to the data stored in the warehouse. 17 / 40

63 Data warehouse refreshment Refreshing a warehouse means propagating updates on source data to the data stored in the warehouse. Follows the same structure as ETL process. 17 / 40

64 Data warehouse refreshment Refreshing a warehouse means propagating updates on source data to the data stored in the warehouse. Follows the same structure as ETL process. Several constraints: accessibility of data sources, size of data, size of data warehouse, frequency of data refreshing, degradation of performance of operational systems. 17 / 40

65 Data warehouse refreshment Detect changes in external data sources: 18 / 40

66 Data warehouse refreshment Detect changes in external data sources: Different monitoring techniques: external and intrusive techniques. 18 / 40

67 Data warehouse refreshment Detect changes in external data sources: Different monitoring techniques: external and intrusive techniques. Snapshot vs. timestamped sources 18 / 40

68 Data warehouse refreshment Detect changes in external data sources: Different monitoring techniques: external and intrusive techniques. Snapshot vs. timestamped sources Queryable, logged, and replicated sources 18 / 40

69 Data warehouse refreshment Detect changes in external data sources: Different monitoring techniques: external and intrusive techniques. Snapshot vs. timestamped sources Queryable, logged, and replicated sources Callback and internal action sources 18 / 40

70 Data warehouse refreshment Detect changes in external data sources: Different monitoring techniques: external and intrusive techniques. Snapshot vs. timestamped sources Queryable, logged, and replicated sources Callback and internal action sources Extract the changes and integrate into the warehouse. 18 / 40

71 Data warehouse refreshment Detect changes in external data sources: Different monitoring techniques: external and intrusive techniques. Snapshot vs. timestamped sources Queryable, logged, and replicated sources Callback and internal action sources Extract the changes and integrate into the warehouse. Update indexes, subaggregates and any other additional data structures. 18 / 40

72 Data warehouse refreshment Detect changes in external data sources: Different monitoring techniques: external and intrusive techniques. Snapshot vs. timestamped sources Queryable, logged, and replicated sources Callback and internal action sources Extract the changes and integrate into the warehouse. Update indexes, subaggregates and any other additional data structures. Types of refreshments: 18 / 40

73 Data warehouse refreshment Detect changes in external data sources: Different monitoring techniques: external and intrusive techniques. Snapshot vs. timestamped sources Queryable, logged, and replicated sources Callback and internal action sources Extract the changes and integrate into the warehouse. Update indexes, subaggregates and any other additional data structures. Types of refreshments: Periodical refreshment (daily or weekly). 18 / 40

74 Data warehouse refreshment Detect changes in external data sources: Different monitoring techniques: external and intrusive techniques. Snapshot vs. timestamped sources Queryable, logged, and replicated sources Callback and internal action sources Extract the changes and integrate into the warehouse. Update indexes, subaggregates and any other additional data structures. Types of refreshments: Periodical refreshment (daily or weekly). Immediate refreshment. 18 / 40

75 Data warehouse refreshment Detect changes in external data sources: Different monitoring techniques: external and intrusive techniques. Snapshot vs. timestamped sources Queryable, logged, and replicated sources Callback and internal action sources Extract the changes and integrate into the warehouse. Update indexes, subaggregates and any other additional data structures. Types of refreshments: Periodical refreshment (daily or weekly). Immediate refreshment. Determined by usage, types of data source, etc. 18 / 40

76 Outline 1 Motivation 2 ETL 3 OLAP Systems 4 Summary 19 / 40

77 OLAP systems The next step is to provide solutions for querying and reporting multidimensional analytical data. 20 / 40

78 Multidimensional cube The proper data model for multidimensional reporting is the multidimensional one. 21 / 40

79 Operations in multidimensional data model Roll up summarize data along a dimension hierarchy. Drill down go from higher level summary to lower level summary or detailed data. Slice and dice corresponds to selection and projection. Pivot reorient cube. Raking, Time functions, etc. 22 / 40

80 Lattice of cuboids Different degrees of summarizations are presented as a lattice of cuboids. Example for dimensions: time, product, location, supplier {all} {time} {product} {location} {supplier} {time, product} {time, location} {time, supplier} {product, location} {product, supplier} {location, supplier} {time, product, location} {time, product, supplier} {time, location, supplier} location, {product, supplier} {time,product,location,supplier} Using this structure, one can easily show roll up and drill down operations. 23 / 40

81 Total number of cuboids For an n-dimensional data cube, the total number of cuboids that can be generated is: n T = (L i + 1), i=1 where L i is the number of levels associated with dimension i (excluding the virtual top level all since generalizing to all is equivalent to the removal of a dimension). For example, if the cube has 10 dimensions and each dimension has 4 levels, the total number of cuboids that can be generated will be: T = 5 10 = 9, / 40

82 Total number of cuboids Example: Consider a simple database with two dimensions: 25 / 40

83 Total number of cuboids Example: Consider a simple database with two dimensions: Columns in Date dimension: day, month, year Columns in Localization dimension: street, city, country. Without any information about hierarchies, the number of all possible group-bys is 25 / 40

84 Total number of cuboids Example: Consider a simple database with two dimensions: Columns in Date dimension: day, month, year Columns in Localization dimension: street, city, country. Without any information about hierarchies, the number of all possible group-bys is 2 6 : 25 / 40

85 Total number of cuboids Example: Consider a simple database with two dimensions: Columns in Date dimension: day, month, year Columns in Localization dimension: street, city, country. Without any information about hierarchies, the number of all possible group-bys is 2 6 : day street month city year country day, month street, city day, year street, country month, year city, country day, month, year street, city, country 25 / 40

86 Total number of cuboids Example: Consider the same relations but with defined hierarchies: 26 / 40

87 Total number of cuboids Example: Consider the same relations but with defined hierarchies: day month year street city country 26 / 40

88 Total number of cuboids Example: Consider the same relations but with defined hierarchies: day month year street city country Many combinations of columns can be excluded, e.g., group by day, year, street, country. The number of group-bys is then 26 / 40

89 Total number of cuboids Example: Consider the same relations but with defined hierarchies: day month year street city country Many combinations of columns can be excluded, e.g., group by day, year, street, country. The number of group-bys is then 4 2 : 26 / 40

90 Total number of cuboids Example: Consider the same relations but with defined hierarchies: day month year street city country Many combinations of columns can be excluded, e.g., group by day, year, street, country. The number of group-bys is then 4 2 : year country month, year city, country day, month, year street, city, country 26 / 40

91 Three types of aggregate functions distributive: count(), sum(), max(), min(), algebraic: ave(), stddev(), holistic: median(), mode(), rank(). 27 / 40

92 OLAP servers Relational OLAP (ROLAP), Multidimensional OLAP (MOLAP), Hybrid OLAP (HOLAP). 28 / 40

93 ROLAP ROLAP servers use a relational or post-relational database management system to store and manage warehouse data. 29 / 40

94 ROLAP ROLAP servers use a relational or post-relational database management system to store and manage warehouse data. ROLAP systems use SQL and its OLAP extensions. 29 / 40

95 ROLAP ROLAP servers use a relational or post-relational database management system to store and manage warehouse data. ROLAP systems use SQL and its OLAP extensions. Optimization techniques: 29 / 40

96 ROLAP ROLAP servers use a relational or post-relational database management system to store and manage warehouse data. ROLAP systems use SQL and its OLAP extensions. Optimization techniques: Denormalization, 29 / 40

97 ROLAP ROLAP servers use a relational or post-relational database management system to store and manage warehouse data. ROLAP systems use SQL and its OLAP extensions. Optimization techniques: Denormalization, Materialized views, 29 / 40

98 ROLAP ROLAP servers use a relational or post-relational database management system to store and manage warehouse data. ROLAP systems use SQL and its OLAP extensions. Optimization techniques: Denormalization, Materialized views, Partitioning, 29 / 40

99 ROLAP ROLAP servers use a relational or post-relational database management system to store and manage warehouse data. ROLAP systems use SQL and its OLAP extensions. Optimization techniques: Denormalization, Materialized views, Partitioning, Joins, 29 / 40

100 ROLAP ROLAP servers use a relational or post-relational database management system to store and manage warehouse data. ROLAP systems use SQL and its OLAP extensions. Optimization techniques: Denormalization, Materialized views, Partitioning, Joins, Indexes (join index, bitmaps), 29 / 40

101 ROLAP ROLAP servers use a relational or post-relational database management system to store and manage warehouse data. ROLAP systems use SQL and its OLAP extensions. Optimization techniques: Denormalization, Materialized views, Partitioning, Joins, Indexes (join index, bitmaps), Query processing. 29 / 40

102 ROLAP Advantages of ROLAP Servers: 30 / 40

103 ROLAP Advantages of ROLAP Servers: Scalable with respect to the number of dimensions, 30 / 40

104 ROLAP Advantages of ROLAP Servers: Scalable with respect to the number of dimensions, Scalable with respect to the size of data, 30 / 40

105 ROLAP Advantages of ROLAP Servers: Scalable with respect to the number of dimensions, Scalable with respect to the size of data, Sparsity is not a problem (fact tables contain only facts), 30 / 40

106 ROLAP Advantages of ROLAP Servers: Scalable with respect to the number of dimensions, Scalable with respect to the size of data, Sparsity is not a problem (fact tables contain only facts), Mature and well-developed technology. 30 / 40

107 ROLAP Advantages of ROLAP Servers: Scalable with respect to the number of dimensions, Scalable with respect to the size of data, Sparsity is not a problem (fact tables contain only facts), Mature and well-developed technology. Disadvantage of ROLAP Servers: 30 / 40

108 ROLAP Advantages of ROLAP Servers: Scalable with respect to the number of dimensions, Scalable with respect to the size of data, Sparsity is not a problem (fact tables contain only facts), Mature and well-developed technology. Disadvantage of ROLAP Servers: Worse performance than MOLAP, 30 / 40

109 ROLAP Advantages of ROLAP Servers: Scalable with respect to the number of dimensions, Scalable with respect to the size of data, Sparsity is not a problem (fact tables contain only facts), Mature and well-developed technology. Disadvantage of ROLAP Servers: Worse performance than MOLAP, Additional data structures and optimization techniques used to improve the performance. 30 / 40

110 MOLAP MOLAP Servers use array-based multidimensional storage engines. 31 / 40

111 MOLAP MOLAP Servers use array-based multidimensional storage engines. Optimization techniques: 31 / 40

112 MOLAP MOLAP Servers use array-based multidimensional storage engines. Optimization techniques: Two-level storage representation: dense cubes are identified and stored as array structures, sparse cubes employ compression techniques, 31 / 40

113 MOLAP MOLAP Servers use array-based multidimensional storage engines. Optimization techniques: Two-level storage representation: dense cubes are identified and stored as array structures, sparse cubes employ compression techniques, Materialized cubes. 31 / 40

114 MOLAP Advantages of MOLAP Servers: 32 / 40

115 MOLAP Advantages of MOLAP Servers: Multidimensional views are directly mapped to data cube array structures efficient access to data, 32 / 40

116 MOLAP Advantages of MOLAP Servers: Multidimensional views are directly mapped to data cube array structures efficient access to data, Can easily store subaggregates. 32 / 40

117 MOLAP Advantages of MOLAP Servers: Multidimensional views are directly mapped to data cube array structures efficient access to data, Can easily store subaggregates. Disadvantages of MOLAP Servers: 32 / 40

118 MOLAP Advantages of MOLAP Servers: Multidimensional views are directly mapped to data cube array structures efficient access to data, Can easily store subaggregates. Disadvantages of MOLAP Servers: Scalability problem in the case of larger number of dimensions, 32 / 40

119 MOLAP Advantages of MOLAP Servers: Multidimensional views are directly mapped to data cube array structures efficient access to data, Can easily store subaggregates. Disadvantages of MOLAP Servers: Scalability problem in the case of larger number of dimensions, Not tailored for sparse data. 32 / 40

120 MOLAP Advantages of MOLAP Servers: Multidimensional views are directly mapped to data cube array structures efficient access to data, Can easily store subaggregates. Disadvantages of MOLAP Servers: Scalability problem in the case of larger number of dimensions, Not tailored for sparse data. Example: 32 / 40

121 MOLAP Advantages of MOLAP Servers: Multidimensional views are directly mapped to data cube array structures efficient access to data, Can easily store subaggregates. Disadvantages of MOLAP Servers: Scalability problem in the case of larger number of dimensions, Not tailored for sparse data. Example: Logical model consists of four dimensions: customer, product, location, and day 32 / 40

122 MOLAP Advantages of MOLAP Servers: Multidimensional views are directly mapped to data cube array structures efficient access to data, Can easily store subaggregates. Disadvantages of MOLAP Servers: Scalability problem in the case of larger number of dimensions, Not tailored for sparse data. Example: Logical model consists of four dimensions: customer, product, location, and day In case of customers, products, locations and days, the data cube will contain cells! 32 / 40

123 MOLAP Advantages of MOLAP Servers: Multidimensional views are directly mapped to data cube array structures efficient access to data, Can easily store subaggregates. Disadvantages of MOLAP Servers: Scalability problem in the case of larger number of dimensions, Not tailored for sparse data. Example: Logical model consists of four dimensions: customer, product, location, and day In case of customers, products, locations and days, the data cube will contain cells! A huge number of cells is empty: a customer is not able to buy all products in all locations / 40

124 HOLAP HOLAP servers are a hybrid approach that combines ROLAP and MOLAP technology. HOLAP benefits from the greater scalability of ROLAP and the faster computation of MOLAP. 33 / 40

125 Multidimensional queries We need an intuitive way of expressing analytical (multidimensional) queries: 34 / 40

126 Multidimensional queries We need an intuitive way of expressing analytical (multidimensional) queries: Operations like roll up, drill down, slice and dice, pivoting, ranking, time and window functions, etc. 34 / 40

127 Multidimensional queries We need an intuitive way of expressing analytical (multidimensional) queries: Operations like roll up, drill down, slice and dice, pivoting, ranking, time and window functions, etc. Two solutions: 34 / 40

128 Multidimensional queries We need an intuitive way of expressing analytical (multidimensional) queries: Operations like roll up, drill down, slice and dice, pivoting, ranking, time and window functions, etc. Two solutions: Extending SQL, or 34 / 40

129 Multidimensional queries We need an intuitive way of expressing analytical (multidimensional) queries: Operations like roll up, drill down, slice and dice, pivoting, ranking, time and window functions, etc. Two solutions: Extending SQL, or Inventing a new language ( MDX), 34 / 40

130 SQL OLAP extensions in SQL: GROUP BY CUBE, GROUP BY ROLLUP, GROUP BY GROUPING SETS, GROUPING and DECODE/CASE OVER and PARTITION BY, RANK. 35 / 40

131 MDX MDX Multidimensional expressions. 36 / 40

132 MDX MDX Multidimensional expressions. For OLAP queries, MDX is an alternative to SQL: 36 / 40

133 MDX MDX Multidimensional expressions. For OLAP queries, MDX is an alternative to SQL: 36 / 40

134 MDX MDX Multidimensional expressions. For OLAP queries, MDX is an alternative to SQL: Academic year Instructor AVG(Grade) 2011/2 Stefanowski /2 S lowiński /3 Stefanowski /3 S lowiński /4 Stefanowski /4 S lowiński /4 Dembczyński / 40

135 MDX MDX Multidimensional expressions. For OLAP queries, MDX is an alternative to SQL: Academic year Instructor AVG(Grade) 2011/2 Stefanowski /2 S lowiński /3 Stefanowski /3 S lowiński /4 Stefanowski /4 S lowiński /4 Dembczyński 4.8 AVG(Grade) Academic year Name 2011/2 2012/3 2013/4 Stefanowski S lowiński Dembczyński / 40

136 MDX MDX query: SELECT {[Time].[1997],[Time].[1998]} ON COLUMNS, {[Measures].[Sales],[Measures].[Cost]} ON ROWS FROM Warehouse WHERE ([Store].[All].[USA]) Seems to be similar to SQL, but in fact it is quite different! 37 / 40

137 Outline 1 Motivation 2 ETL 3 OLAP Systems 4 Summary 38 / 40

138 Summary ETL process is a strategic element of data warehousing. Main concepts: extraction, transformation and integration, load, data warehouse refreshment and metadata. New emerging technology... OLAP systems: ROLAP, MOLAP and HOLAP. 39 / 40

139 Bibliography J. Han and M. Kamber. Data Mining: Concepts and Techniques (second edition). Morgan Kaufmann Publishers, / 40