The appeal of analytics is driven by “just the proverbial infinite scale of the cloud and movements in platforms aggregating and concatenating large datasets for these purposes,” said Joe Corvaia, vice president of cloud computing for Evolve IP. “The cloud is the underlying movement that has enabled this type of innovation, scale, and capacity to house these datasets in a meaningful way, coupled with the sheer resources to process and mine the data to extract value, and then make them available to companies of all sizes and budgets.” Not only has cloud boosted the adoption of analytics, but it will be the primary contributor to its acceleration over the years to come, he noted.
This becomes especially important as data becomes a business unto itself for many organizations. “Companies—small, medium, and large—do not have a shortage of data, but they are challenged with how to harvest that data—how to turn data into information,” said Mark Wojtasiak, segment marketing manager for cloud at Seagate. “Information enables them to gain a competitive advantage, and this is something more and more companies are willing to subscribe to as a service.”
Fortunately, there have also been impressive advances to solve some of these new integration challenges, said Fisher. “Big data stores are now, more than ever, accessible via standard and well-known interfaces like SQL. Data services are providing API interfaces to help organizations access and publish data with documented access support. This provides organizations with the ability to get and provide access to data that was otherwise difficult, expensive, or impossible access.”
Cloud Databases Can Add Complexity to Data Environments
The fact that cloud databases will need to integrate, or at least co-exist, with existing on-premises data environments will add to the complexity of today’s data environments, many observers feel. “IT adds new technologies and rarely moves off old technologies completely,” said Mike Madden, general manager of mainframe for CA Technologies.“For example, it’s unlikely that our large mainframe customers will move or migrate their large mainframe databases to the cloud. But it is very likely that they might use the cloud for data analytics, and therefore utilize Hadoop or another database for the purpose. And, they will be doing ETL operations to move data back and forth between the old and the new. This will bring about security and other operational issues that didn’t exist before.”
It all boils down to architecture. An architecture designed first around clustering, replication, and ease of virtualization is critical to cloud deployments, said Ryan Betts, CTO of VoltDB. “This means eliminating expensive shared storage with modern shared-nothing cluster architectures.” In the last several years, there has been an explosion of data management solutions—and with good cause, he said. “The legacy architectures of incumbent products do not meet the elastic, shared-nothing, virtualization, and horizontal scaling requirement for cloud deployments.”
Legacy applications are less likely to go to the cloud, agreed John Hawkins, senior director of consulting services at RiverMeadow Software. “The reason for this the years of investment in on-premise database models, tools, and, most importantly, data architects, developers, and administrators trained in these on-premise models. These folks have invested years using on-premise data models. It’s difficult to change how you do things overnight; business must go on. It can be challenging to adopt a new model built on platforms that are cloud-ready with the resources and tools all geared toward on-premise models.”
There’s no consensus on whether moving to the cloud will help to reduce the complexity of database management. It depends on the type of organization. For example, smaller organizations or startups benefit from the relative simplicity of databases in the cloud. “Most times, when a paradigm shift occurs, things get simpler for greenfield projects and more complex for those with existing systems,” said Pasqua. “Cloud is no different. For example, a startup may be cloud native from day-one. It can benefit from the reduced management burden right away.”
As a result, big data is driving a great deal of interest in cloud databases, especially as real-time analytics and the Internet of Things takes hold. “A traditional on-premise infrastructure just won’t work—imagine the poor IT admin who has to specify, order, and install and configure hardware as the data grows by terabytes every month,” said Dev Patel, CEO and founder of BitYota. Thus, the cloud becomes the “logical and compelling place to cost-effectively handle access to and store this exploding volume of data,” he added.
Big data analytics lends itself to cloud deployment naturally, added Anupam Manglik, vice president, cloud services and application migration and development for NTT Data. “Big data requires that the underlying infrastructure scales as the data velocity and volume grows. Cloud is a natural destination for big data applications. Cloud also facilitates creation of big data proofs of concept by enabling enterprises to provision resources on-demand to prove out big data use cases.”