Real-time BI: High data availability for users

Loading data warehouses

Real-Time4 Business Intelligence can load an active data warehouse without triggering production and based on changed data only, resulting in a significant reduction in the volumes exchanged.

By propagating data synchronously or asynchronously between databases, this solution can synchronize targeted data in a central data warehouse with local or remote data marts.

Loading an ODS in real time

Real-Time4 Business Intelligence can load an Operational Data Store (ODS) in real time from several different sources. The ODS is subsequently queried by a BI tool (SAP BusinessObjects, IBM Cognos, Oracle Hyperion, etc.). This type of architecture makes the implementation of high value-added projects such as MDM (Master Data Management) or data quality much easier.

Real-Time4 Business Intelligence was successfully implemented in the information system of a large transport and logistics company to load the ODS in real time. As soon as a transport request is entered on the main administrative site running an AS/400 system, it must impact the 50 Oracle databases used by the various warehouses in less than 5 seconds, in order for transport operators to immediately process the request. Real-Time4 Business Intelligence was compared with MQSeries and SonicMQ-based solutions. Its rapid deployment (less than a 4-day setup) and fast response time (less than 300 ms) were the main reasons behind the customer's choice. It took 4 days to configurate the architecture parameters and replication rules.

Loading large volumes of data

Real-Time4 Business Intelligence is used to manage large volumes of data upstream from the loading of the BI system.

Large volumes and 24/7 operation are the two main characteristics of this major luxury mail order company. Despite the constraints (the ODS needed to be loaded as an "airtight" entrance to the data warehouse without causing delays or disrupting production, which would otherwise mean a risk of lost sales and customers), the outcome was a success: Gamma Soft's technology can easily accept up to 400,000,000 transactions on peak days. Its transformation scripts effectively optimized the data flows exchanged. 

Loading BI systems: summary

To meet the users' need for high data availability, Real-Time4 Business Intelligence delivers reliable, easy-to-implement solutions:

• Real-time data exchange to have information without needing to wait for the next day.
• Extraction processes on source applications with a very low CPU impact, often below 1% of CPU consumption.
• Time windows for extracting data are discarded: Real-Time4 Business Intelligence is the recommended real-time BI solution for 24/7 systems.
• Data loaded on the fly to easily obtain data from tables whose latest records cannot be identified: reading the elements in the transactions log serves to identify only those items that have changed. As a result, only information that has changed is conveyed.
• Extremely high fault tolerance (network, system) with automatic disaster recovery processes.
• Flows between sources and targets that are secure and reliable.
• An optimized infrastructure: when the BI system is hosted on a powerful server, using Real-Time4 Business Intelligence in an ETL-type architecture is recommended. Real-Time4 Business Intelligence delivers the data to the target system which then performs the transformation process by exploiting the system's full power. As a result, there is simply no need to assign a powerful intermediate machine for performing transformations on the data.
• A tangible contribution to the company's cost reduction policy: Real-Time4 Business Intelligence helps to reduce infrastructure and database licensing costs by enabling the migration of databases and applications to other less costly tools and systems (Open Source).