In this "data explosion" era, data has become a key production factor for all industries. Traditional data integration technologies are facing stiff challenges. Applications are fragmented. Data is disconnected and stored in silos. Data security is managed in a way that hampers data integration and access across business units.

Under this circumstance, enterprises use ETL tools to extract, transform and load data from multiple data sources to obtain business data, which raises various issues such as huge integration workloads and high security risks .

To address those issues, RapidsDB Data Virtualization Management (DVM) is built on our ingenious next-generation of in-memory database technology, independently developed by the RapidsDB team. RapidsDB DVM can mask the format, geographical location, system, and other attributes of data. It provides a unified domain-wide SQL data interface layer to users.

RapidsDB DVM controls access to the underlying data stores across different data sources via the dynamic and pluggable connectors. The connectors manage the metadata of the objects (usually tables or files) in remote data stores and provide it to the embedded query execution engine as the ANSI-SQL based schemas. This mechanism enables users to treat the objects of the entire data source set as a single, unified SQL database.

Through the ANSI-standard SQL interface layer offered by RapidsDB DVM, users can quickly access all the data of an enterprise, which includes data stored in traditional databases, big data sources, cloud storage, and IoT systems. Data can be accessed, combined, transformed, and delivered at a lightning-fast speed in a cost-effective way. With this solution, complex data service use cases of various industries can be handled effectively.

Example: Query MySQL+Oracle data with one SQL statement based on RapidsDB DVM: Step 1: Connect to MYSQL and prepare data via RapidsDB DVM >use connector MYSQL_TEST; >CREATE TABLE Orders (OrderID INT UNSIGNED NOT NULL DEFAULT 0, CustomerID INT UNSIGNED NOT NULL DEFAULT 0, OrderDate DATE); >insert into Orders (OrderID, CustomerID, OrderDate) VALUES (10308, 2, '1996-09-18'); >insert into Orders (OrderID, CustomerID, OrderDate) VALUES (10309, 37, '1996-09-19');

Step 2: Connect to Oracle and prepare data via RapidsDB DVM >use connector Oracle_TEST; >CREATE TABLE Customers (CustomerID INT UNSIGNED NOT NULL DEFAULT 0, CustomerName VARCHAR(50) NOT NULL, ContactName VARCHAR(50) NOT NULL,);

insert into Customers (CustomerID, CustomerName, ContactName) VALUES (1, 'Alfreds', 'Maria Anders'); insert into Customers (CustomerID, CustomerName, ContactName) VALUES (2, 'Ana', 'Ana Trujillo');

Step 3: Query MySQL+Oracle data with one statement via RapidsDB DVM

select o.CustomerID,o.CustomerName,c.CustomerID from Oracle_TEST.demo.Customers o join MYSQL_TEST.sqltest.Orders c on o.CustomerID=c.CustomerID;

Customer Case An AI microservice provider A distributed data analytics engine and cloud data warehouse solution was adopted by the original system, which had led to the following pain points: (1)The dedicated database created data silos, increasing the total cost of ownership (TCO); (2)The complex technology stack incurred high development, O&M and training costs for the operational staff; (3)As the data volume continued to grow, customers demanded higher database performance. , However, the overall performance was not ideally improved with the cluster expansion.

Highlights of RapidsDB DVM: (1)It -supports real-time queries, especially for geospatial data and text documents; (2)The data virtualization capability enhances collaboration between application developers and business users; (3)Query results are returned within seconds, dramatically improving user experience and team collaboration.