Abstract: Recently, FanRuan officially released the big data direct connection engine FineDirect module . Through this module, enterprises can directly connect to existing data sources on the basis of applying the original functions of FineBI. Whether it is a traditional relational database, or the Hadoop ecosystem or Mpp architecture, they can directly fetch and analyze data by themselves.
At present, the application of data by enterprises is still dominated by the combination of data warehouse and BI, on the other hand, more and more enterprises have gradually introduced big data computing platforms. Personalized solutions, growing data, and increasing demands on BI tools.
Gartner also pointed out in its 2017 BI report: In the next five years, Hadoop/Spark-based, search and visualization-based data exploration and analysis functions will be integrated into next-generation data exploration and analysis products as components of new BI and analysis platforms.
In the past, FineBI has been using the FineIndex (original cube) method to fetch and analyze data, which is the method of "database-FineIndex-front-end analysis". The FineIndex here is equivalent to an intermediate multidimensional database for storing data tables and escaping data associations, which greatly improves the efficiency of subsequent front-end analysis and processing of data. Because of direct sql fetching, the efficiency is limited by the database itself. When the amount of data is large, general analysis tools are easily stuck or even memory overflows cause the system to become unresponsive. This is also the original intention of the FineIndex solution. The existence of FineIndex has two meanings, one is to improve efficiency, and the other is to perform secondary integration processing of data.
Today, FineBI has the FineIndex engine (original cube) and the new FineDirect direct connection engine, which can be used together to meet different scenarios. Enterprises can prepare two types of data according to their actual needs. The FineIndex mode is used to configure the data that is not frequently updated and has low real-time requirements; the FineDirect engine is used to configure the data with large data volume and real-time analysis requirements.
FineDirect mainly deals with real-time analysis of data and analysis and processing of large data volumes. For example, financial industry transaction risk analysis, real-time analysis of each transaction. In addition, many enterprises already have their own big data computing platforms, such as hadoop, kylin, greenplum, vertica, etc. The FineBI direct connection engine provides the function of connecting to these data platforms.
At present, after many internal tests and actual application scenarios, the FineBI direct connection solution has had successful cases in many companies:
Case 1: SAIC Motor Group - Real-time Analysis of Large Volume of Data
Each car produced by SAIC Group will send back the GPS positioning data generated when it is driving on the road every 5 minutes. At present, the accumulated amount of data is about 500 million. The FineDirect data engine helps users to solve the display of these massive data. The problem can be displayed within 5S.
Case 2: Bank of Jiangsu - Cross-Data Source Analysis
The processing capability of FineDirect across data sources, the unified organization table of the row is associated with the data tables of multiple different systems, and there is no need to maintain a separate table in each system that uses the organization table, which ensures the integrity of the data. consistency. And with the help of the parameter function of FineDirect, the index of the existing Oracle database of the business is fully utilized to realize more flexible real-time analysis.
Case 3: Beibei Bear - Real-time Analysis of HANA Database
The company rebuilt the hana database in early 2017 to meet the needs of real-time big data analysis, and used FineBI's FineDirect direct connection engine to connect to the hana database, solving the long-standing problem of real-time big data analysis. Real-time analysis of supplier purchases, real-time comparison of free sales dates, real-time analysis of member sales and other scenarios have been practically applied.
technical details
1. 20 kinds of big data platform docking
Supports docking of up to 20 types of big data platforms such as hadoop, vertica, sap hana, greenplum, kylin, etc., to maximize the computing advantages of the platform itself.
2. Fully visualized configuration
Visual SQL engine, visual multi-dimensional analysis, from data configuration to data presentation. Covers two stages of self-service BI: self-service data retrieval and self-analysis.
3. Intelligent caching mechanism
The intelligent caching mechanism maximizes the saving of computing resources, ensures the availability of the database under the pressure of high concurrency, makes the data acquisition speed better, and saves resources.
4. Flexible parameter application
灵活易用的参数配置功能，可视化参数设置界面，支持通过参数来实现更加灵活的数据获取。
5、敏捷易用的数据模型
可视化的数据关联模型配置，智能化模型配置推荐，跨数据源关联消除孤岛，内存化与库化机制自助可选，满足企业绝大多数的实际应用场景。
6、多级权限控制
支持多级权限控制，满足体量更大的集团用户的权限分配和下放。
7、多数据源覆盖
支持通过帆软报表FineReport的服务器数据集连接功能，实现更多数据源直连覆盖。
8、与FineIndex搭配使用
支持与FineIndex同平台应用，企业可以根据自身的应用场景灵活的选择两种模式，满足不同需求。