Data Virtualization: Federated and Hybrid

Posted by Krishnamoorthy on Oracle Blogs See other posts from Oracle Blogs or by Krishnamoorthy
Published on Tue, 20 Nov 2012 09:36:58 +0000 Indexed on 2012/11/20 11:15 UTC
Read the original article Hit count: 283

Filed under:

Data becomes useful when it can be leveraged at the right time. Not only enterprises application stores operate on large volume, velocity and variety of data. Mobile and social computing are in the need of operating in foresaid data.


Replicating and transferring large swaths of data is one challenge faced in the field of data integration. However, smaller chunks of data aggregated from a variety of sources presents and even more interesting challenge in the industry.


Over the past few decades, technology trends focused on best user experience, operating systems, high performance computing, high performance web sites, analysis of warehouse data, service oriented architecture, social computing, cloud computing, and big data.


Operating on the ‘dark data’ becomes mandatory in the future technology trend, although, no solution can make dark data useful data in a single day. Useful data can be quantified by the facts of contextual, personalized and on time delivery.


In most cases, data from a single source may not be complete the picture. Data has to be combined and computed from various sources, where data may be captured as hybrid data, meaning the combination of structured and unstructured data. Since related data is often found across disparate sources, effectively integrating these sources determines how useful this data ultimately becomes.


Technology trends in 2013 are expected to focus on big data and private cloud. Consumers are not merely interested in where data is located or how data is retrieved and computed. Consumers are interested in how quick and how the data can be leveraged.


In many cases, data virtualization is the right solution, and is expected to play a foundational role for SOA, Cloud integration, and Big Data.


The Oracle Data Integration portfolio includes a data virtualization product called ODSI (Oracle Data Service Integrator). Unlike other data virtualization solutions, ODSI can perform both read and write operations on federated/hybrid data (RDBMS, Webservices,  delimited file and XML). The ODSI Engine is built on XQuery, hence ODSI user can perform computations on data either using XQuery or SQL.


Built in data and query caching features, which reduces latency in repetitive calls.


Rightly positioning ODSI, can results in a highly scalable model, reducing spend on additional hardware infrastructure.

© Oracle Blogs or respective owner

Related posts about /Oracle