BI tools currently on the market are mentioned agile BI solutions. Agile BI solutions automate technical support provided by the program are mainly from the data source to get the number of BI front-end tools show. This agile BI solutions in the enterprise data volume is not very large, the support is very good run. PowerBI can support a large number of data processing, but the hardware requirements are very high. But the huge amount of data becomes more and more will lead to BI report appears to run slow, big-screen data appear to show a delay so the phenomenon.
If the project is small, such as dynamic sales report image above, just to show the amount of data in EXCEL. Use PowerBI tools fully meet everyone's needs. To really do a whole number of positions agile BI solutions, or need to add a data warehouse in the middle of data sources and BI front-end tools. Processing data in the data warehouse without any invasive source for data, nor will it affect the source data systems. It may build a data warehouse tools used SSDT should know that to build a data warehouse is still very complicated. Build a data warehouse or data warehouse need the help of automated tools.
Data warehouse automation tools are becoming more and more mainstream, and now they are obvious benefits:
1, fast delivery
2, lower development costs
3, short development cycle for a business intelligence project is completed, no longer need to wait three to six months.
4, low maintenance costs, without having to invest a lot of maintenance technicians
There are already some ETL tool automation technology, I am in front of the ETL process and ETL tool introduced already mentioned here will not do too much description. There are some things you should know to assess the dimensions of products from different vendors.
A low degree of automation tools that some of the data warehouse.
Some data warehouse automation tools there can not be automated process model, most developers are doing is using the example of a simple star schema report. This is a simple data source, but when you need to integrate data from multiple data sources, things get complicated. Some tools in conjunction with you before the adoption of an intricate process, to upgrade data generation star schema. It is not very automated. Data warehouse automation tools should be able to handle 70% of the work, without the need for additional data modeling or ETL programming. Unless there is customer demand for customized, re-modeling and programming.
Second, some tools require a lot of consulting work to achieve results.
Ask your data warehouse automation software vendor this simple question: "How many of you implement BI project implementation consultants, the implementation cycle needs?" A robust BI project implementation team, if you have done most of automation tools deal with. The actual supplier of automation technology has matured, they are doing more sort, index data fit the model library client companies reporting metrics. More work is in demand models confirm preliminary research stage, the implementation of the deployment of BI projects is actually very fast.
Third, the snowflake model and star model
If you are planning to build their own data warehouse, then you choose the data warehouse automation tool should automatically perform some or all of any of the above. Some of the data warehouse automation tools still require you to manually target model design and use their own tools to fill it. It's not automated, you might as well go back to using ETL tools. A good data warehouse automation tools to automate the design code model and populate it. Lets you choose between snowflake model and star-shaped mold.
Fourth, the data warehouse target database.
Many data warehouse automation tools limit you to just one target database platform, while others will let you create a data warehouse more. You may wish, in the future, to move to a different database platforms (for example, from S QL Server to Oracle, or from S QL Server to another edition of S QL Server), so you may need a data warehouse automation tool that provides a future migration options for you.
Five-dependent scheduling
For any data warehouse project data is needed at a particular time, and loading in a certain order. For example, when data from multiple sources is combined, you may want to have all the add before, first you can start building your data in the table, and they must be updated before you can update report your star schema. An enterprise-class data warehouse automation tools to understand these dependencies, automate and automatically complete all the necessary processes and run them in the correct order.
Sixth, data warehouse automation tool functionality
Data warehouse automation tools including system management, business management bus, dimensional model management, job management, application management, metadata management, data management and industry standard library modules indicators. Personally I think that really should have robust software features on the map or achieve 34 subsystems dimensional modeling toolbox mentioned.
System Management: This module is mainly management information systems, including the various sources of information systems, data warehouse systems, job scheduling system, report viewer system. Import metadata information may be automatically source system by the source system, provides dimensional modeling data and inspection data structure changes.
Bus Business Management: Each source system has multiple business processes, each business process will involve multiple dimensions entity. Business processes and dimension entities bus architecture, a unified management system for each source of business bus.
Dimensional Model Management: physical model and the mapping between design based on naming conventions, to ensure the unity named. Can be generated automatically build a table script, the script is performed automatically in the data warehouse.
Job Management: ETL program automatically generates scripts and programs based on the query templates automatically generate the dependencies of the job. Program code to ensure unified and standardized.
Application Management: unified management reporting, interface to business metadata information.
Standard data management: unified management code naming conventions, data types are mapped, used the word library, index definitions, ETL template.
Metadata Management: In the development process, various metadata stored information can be checked against the code specification metadata, impact analysis data, monitoring data anomalies.
Document Management Project: Memo information can be generated automatically when the data processing BI project, converting field may know be described in detail, extraction rules, etc.
Industry index Library: adapt customer data accumulated by the company's industry data indicators, low-cost delivery of agile BI project.
Seven summary
ETL tool introduces delays and the risk of a lot of time into your business intelligence projects. Worse, they look forward to your business users can learn to master data warehouse tables and fields, do not understand the real business data warehouse user, so the inevitable change will take a long time to resolve. Really allows business users to understand business terms only, using the semantic model is the best way to solve this kind of problem. You can use agile ETL tools, to complete the design of the model. If you estimate the traditional data warehouse project, we need half a year, in the same project with good height and configuration of the data warehouse automation tools to complete the deal in just a few weeks. Time for any company and individual is very important, so is the selection of data warehousing tools need to be very careful.