As enterprise applications becoming increasingly distributed as a result of broader cloud adoption, so too will enterprise data stores become increasingly distributed. While this presents general data management issues, of particular importance is the challenge of data integration in the cloud. Though ETL type processing continues to be a viable method of integrating data, the siloed nature of cloud application deployments radically increases the number of transactions required to consistently aggregate and integrate enterprise data. Add to this the fact that most existing ETL tools are optimized to work between relational databases, while many cloud solutions are based on NoSQL structures and may be required to migrate data through intermediary ?staging? platforms. As transaction volume increases exponentially, so does bandwidth usage and bandwidth cost and so CIO's need to work hard to control this spiral while facilitated clean integration.
Takeaways: