With the concern over data security in the cloud lessening, businesses are opening up to new ways of managing and providing access to corporate information. In the process, business and IT executives alike are getting more comfortable with the idea of processing and storing data in the cloud, with comfort growing “as they realize the benefits,” said Sumit Nijhawan, CEO and president at Infogix. Such benefits span the range of critical data management touchpoints, including “faster deployment by eliminating hardware procurement and commissioning, reducing required IT staff, accelerating deployment, and business continuity.”
Data as a Service—at Your Service
Rising adoption of and increasing comfort levels with cloud-based data are fueling a new treatment of data across enterprises —data as a service, or DaaS. “The concept is similar to software as a service, in that the data platform or service is provided on demand,” said Mita Mahadevan, group engineering manager, data engineering and analytics for Intuit. “The user is typically a developer or a data scientist that accesses the tools but remains agnostic to the specifics of the enabling platform and infrastructure. Some services, for example, crawl the web for relevant data and provide analytics, while others provide platforms that enable companies to analyze product usage patterns.”
Within Intuit, DaaS is now offered enterprisewide, Mahadevan continues. “Our data engineering and analytics division offers data services for use by employees across the company to create user profiles, conduct A/B tests, deliver personalization and more.”
However, industry observers are divided on whether databases themselves should or will be in the cloud, which is often referred to as “database as a service,” or DBaaS. Sometimes, it’s best to simply leave well-enough alone, McNabb pointed out. “For on-premises production databases that are meeting the needs of the organization very well, why introduce network latency or a change in licenses?” he asked. “There is not much value in moving an on-premises production database to the cloud.” Still, for new implementations or applications, it’s worth looking at cloud-based database options, McNabb added. “Net-new databases are a different story. With these, putting a database in the cloud for a new usage or using Hadoop or Amazon Redshift to process large volumes of data makes sense.”
Bob Muglia, CEO of Snowflake Computing, observed that he has seen substantial interest in DBaaS for a range of purposes. “When it comes to OLTP databases, many people have already run those databases themselves in the cloud and new offerings from the major cloud vendors are seeing rapid growth,” he said. “Although slower to emerge, data warehousing in the cloud is also becoming increasingly attractive as demonstrated by the rapid growth in adoption of new cloud data warehousing options.”
Taking the as-a-service paradigm to its next logical step, big data as a service, or BDaaS, also is seen as a way to deliver analytical capability widely across enterprises. “It is still early in its adoption cycle, but we’ve seen a significant rise in acceptance in 2015 with hosted Hadoop services such as Microsoft’s HDInsights, Amazon’s Elastic MapReduce, or fully managed big data services such as Altiscale,” said Curtis. “In any of these services, efficient data ingestion and integration of data from multiple sources for processing with Hadoop are essential considerations when making BDaaS work for something other than sandboxes or testing using batches or snapshots of data.”