Data architecture and cloud computing were not a natural match at first, according to TechTarget reporters taking part in the latest Talking Data podcast. That was because concerns about security and the sheer effort to move accumulated hordes of data to the cloud kept many IT data-oriented operations on the sidelines. The fact is, for some risk-adverse individuals, the sidelines were the most comfortable place to be.
But that is changing, TechTarget reporter Jack Vaughan told colleague Ed Burns. Now data and cloud are going forward as if hand in hand. As more data professionals take to the cloud, the very nature of data architecture could change as well.
Vaughan said research for recent SearchDataManagement.com stories shows that many years of Web applications have in some cases shifted the balance of data to the cloud. That is partly because so much data is now actually created on the cloud or the Web in the first place.
Mobile devices, too, spawn new data needs. Perhaps the most pressing change of all is that Web and mobile clients alike create an ever-larger pool of applications users. As conveyed in recent SearchDataManagement stories, the new environments are leading to changes in some basic premises of data handling.
Traditional SQL-based relational databases were oriented toward single-server deployment and disk-drive storage. That approach is tested in cloud settings -- leaving an opening for more ''cloud-friendly'' architectures like NoSQL and NewSQL.
In the podcast, the reporters discuss use cases that include a system that must meet highly diverse Web customer reporting requirements and an application performance monitoring service using NoSQL to collect real-time operations from more than 800 million mobile devices.
The participants say that build-or-buy decisions are now more often accompanied by consideration of the rent option as data and cloud going forward. With IT budgets stretched, and a trend toward Web data firmly in place, distributed cloud data schemes are likely to gain more traction.