KShare

Data Sharing Service Platform

Currently, the number of application systems in various organizations is growing explosively, from a few to dozens or even hundreds. Data exchange between application systems is becoming more and more urgent, but also more and more complex and difficult. In order to achieve interoperability between applications, the current implementation methods are mainly divided into the following categories:

✔Database point-to-point direct connection (such as DB-LINK)

✔Intermediate library or table

✔Customized interface development

✔ESB Enterprise Service Bus

The above methods are either simple to implement but difficult to match complex business processes; or the technology is too complex and highly invasive to the business, requiring major changes on the application side, resulting in huge implementation costs.

When there are only a few or a dozen databases in an organization, it is feasible to centrally synchronize the databases within the enterprise through ETL or data lake technology. However, when the number of databases reaches dozens or even more, it becomes extremely challenging to synchronize the data to a centralized database through ETL or data synchronization. In this case, if the data is still centralized through data synchronization or ETL, the operation and maintenance cost and synchronization failure rate will increase significantly, which will ultimately affect the data quality of the centralized database and inconsistency with the production database.

At present, in the fields of medical care, universities, finance, etc., the number of application systems and databases is as high as hundreds. In order to achieve the interconnection between applications and the central database or ODS layer, various organizations have invested huge manpower and financial resources, but the effect is not ideal. In this context, the concept of Data Fabric has emerged. Its core idea is to access the production data of various business systems in a centralized and real-time manner in a unified and real-time manner. This data management concept that does not rely on data movement and synchronization greatly simplifies the data management process and improves data accessibility, consistency, and security.

KShare data sharing service platform is built based on the concept of Data Fabric, aiming to solve the thorny problem of data sharing between the growing business systems in organizations now and in the future, and to separate data sharing from various application systems and database systems, making it an IT infrastructure and core platform system.

Through the ODBC driver, multiple data sources are connected in real time to form a virtual centralized database system. Users can use a SQL statement to associate tables in multiple different production libraries in the virtual data lake to achieve complex cross-library queries. In addition, through the real-time access interface of the virtual data lake, users can directly generate the central library and ODS layer through SQL, thereby greatly reducing the complexity of ETL processing.

Users can publish interfaces by filling in SQL or anonymous blocks to achieve real-time data sharing and exchange between different application systems, avoiding tedious interface development work and point-to-point database direct connection. At the same time, it can also make up for the shortcoming of ESB enterprise service bus that is not real-time.

Based on the virtual data lake, KShare can perform dynamic and real-time desensitization for third-party applications and RESTful API services to avoid security risks such as data leakage.