Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Operational transformation wikipedia , lookup
Data Protection Act, 2012 wikipedia , lookup
Expense and cost recovery system (ECRS) wikipedia , lookup
Data center wikipedia , lookup
Data analysis wikipedia , lookup
Forecasting wikipedia , lookup
Database model wikipedia , lookup
3D optical data storage wikipedia , lookup
Data vault modeling wikipedia , lookup
OLAP Reporting Tools – Are you really ready for one! Justify (By Shanker Jha) Now a days olap reporting is getting very popular due to excellent reporting capabilities like fast response, fast navigation across dimensions and better presentation of data for decision making etc. The tools are very well advertised followed by the demonstration of the simulations/proof of concept by the vendors. Some times satisfied with the demonstrations and sometimes in hurry to grab one, users deem the tools useful for them. Going for an olap reporting tool is not always a wise decision. The cost of the tools does not justify the needs of the small organizations where limited transaction processing takes place and mostly ad hoc transactional reporting is in demand. The tools are good for the organizations where heavy data processing is done and analytical processing is required. In general when an organization goes for an olap Reporting tool it may or may not have a data warehouse in place. However tool vendors are okay with any thing. Let us analyze a case when an Organization is not having a data warehouse in place and it is going for an olap reporting tool. *The data of the organization happens to be in relational model and most of the olap reporting tools talk about data warehousing terms like facts,dimensions etc. *There are limitations with the OLAP Reporting Tools on the use of data Sources/multiple data sources or heterogeneous data sources. * The representation of same information among the different data sources may vary .for example if a data is coming from two ERP system the same Item may be represented by two item codes like X123AZ in one system and E-X123AZ## in other ERP system. So to consolidate the data, Data cleaning/transformation will be needed or the organization will end up loosing data * Data transformation will be needed for incompatible data type representations for the same field among different data sources. For example: sales price is captured in one system as Number -decimal 2 places, in another system as Double-decimal 5 places. You will always need to be extra careful at the time of designing the cube other wise loading/calculating a cube may fail or cube may be improperly loaded. * Most of the OLAP reporting tools do not provide efficient Extraction, Transformation or Loading facilities. For the proper ETL we are expected to have an ETL tool. So if the organization uses multiple/heterogeneous data sources and needs ETL before fulfilling the reporting requirements we will have to go for an ETL tool or we may not be able to utilize all the Data for the reporting purpose, i.e we may end up in loss of Information. * Storage for extracted and transformed Data will be needed before we load the data in the cubes. * Designing one cube say a Sales data Cube, may end up referring 10-15 or more tables. It will take item number from sales table and corresponding description, Item type code from item master table. As the cube is intended for analytical reporting it is expected to facilitate drill up, down and across the data on different hierarchies so item class description, item section, item session etc will be needed. For all these data the respective tables will be required to be accessed. * At the time of Cube processing data will be extracted from sources satisfying all the Conditions specified (Of courses a hell lot of joins), do the calculations and store the Data (seems simple☺). At the same time another cube processing is needed which references 6 of the tables referred in sales data cube. The same may go for 5, 6, 7….. Any number of cubes. This all will consume a lot of processing time. And all these processing will actively involve the production environment i.e. our OLTP system. So there will be limitations with * The time and frequency of the processing of the cube as nobody wants transactions to be affected. * The number of the cubes and data in the cubes. * Benefit of ETL: In our arena, technology changes are very fast and new things keep coming so we can expect new reporting tool more efficient than the present in near future, at that time we won’t be able to get the benefit of all the ETL we did to populate the cubes now. *Multiplicity of Processing: Acquisitions are very common these days. Acquired companies may be using their OLAP tool and transition to new one may be expensive to achieve. In that case same processing may be done twice for both the tools. * Historical Information: The historical data and slowly changing dimensions can not be addressed directly non datawarhouse environment. So before going to the OLAP reporting tool or Data mining tool it is suggested that organizations should go for a proper data warehouse implementation. Some of the benefits, which can be observed while a Data warehouse is in place before going for a OLAP reporting tool, are followed. * Variety of choices for OLAP reporting tool which may be restricted if one is not having a data modeled in multidimensional model. *The datawarehouse can be populated once and after that cubes can be processed till next ETL session without affecting the Transaction. *The cube design also becomes easier if the data is already modeled in the multidimensional model in the datawarehouse. *Processing time of the cube also gets reduced due to less number of joins now, as data for the cubes will be fetched from the datawarehouse based on the multidimensional data model. *We can address many data related issues at datawarehouse level and concentrate on better OLAP report design. *The data Extracted, transformed and loaded today will be available for the future. *Migration to new tools, uses of multiple OLAP reporting tools or Data Mining tools are always possible. Some OLAP tools also provide the facility for transactional report as well. There are some transactional reports which are required to be generated frequently, and there are some reports which are generated once in a day against the transactional system, the number of these once in a day reports are also high. Once in a day transactional report can be generated against datawarehouse if a data warehouse or relevant data mart is updated regularly.Sometimes users may need to run a Once in a day transactional report on adhoc basis ,In such cases we can have this provision for power/super users and normal ‘Once in a day ‘ report for others. Having a datawarehouse in a place always helps in having an optimal solution for analytical reporting needs. Shanker Jha can be reached at [email protected] or [email protected] He is presently working with CGI – Mumbai having 3+ yrs of experience in datawarehousing.