Emerging Alternatives for Data Management

Bookmark and Share

The staid relational database has been under attack from all quarters—it’s too inflexible, it’s too siloed, it’s too slow for today’s real-time requirements. However, relational databases still power many of the most critical data environments across enterprises—both core legacy systems as well as new-age systems of engagement. The technology typically serves as the foundation for transaction processing systems, including ecommerce websites, enterprise applications such as ERP and CRM systems, business analytics, and many other applications. There are plenty of good things that can be said for relational databases—they are mature and stable; the query-language SQL is feature-rich and versatile; and most major programming languages support them.

In today’s enterprises, then, the question becomes: How are new technologies—especially the cloud and NoSQL databases—affecting relational database adoption? That was the key question behind a new survey of data managers and professionals, conducted by Unisphere Research, a division of Information Today, Inc. The research shows that relational databases continue to hold their place as both the biggest source of transactional data in most organizations and the biggest source of growth for transactional data.  Nearly three quarters (74%) of the respondents said relational databases generate the most transactional data in their organizations. More than half (54%) indicated that relational databases are the fastest growing source of transactional data.

Accordingly, while data is expanding from a rapidly multiplying range of sources, the prime source of enterprise transactional data continues to be relational data. Almost 55% of the respondents said that relational databases were the biggest source of transactional data growth while almost 43% pinpointed enterprise applications as the major source of growth.

The survey explored the movement of databases to the cloud. Over the next 3 years, the cloud will become an increasingly important data repository site, with 60% of the respondents indicating they plan to store more data in the cloud than on-premises.

There are a number of reasons for deploying databases to the cloud. The primary motivating factor to moving a database to the cloud is scalability, which was tabbed by more than 70% of the respondents. Clearly, the relentless growth of data has put scalability of the data infrastructure on the top of many IT departments’ agenda and the scalability of cloud implementations is apparently its most attractive feature, by far. The second most enticing aspect of the cloud for those who have already started to use it is overall cost saving, followed by a range of different opportunities such as utilization of resources and cost efficiency.

In addition, companies with databases in the cloud anticipate that the amount of data stored there will grow at about the same rate as the overall amount of data under management generally. As a result, over time, the cloud will become an increasingly important element in the data infrastructure.

The survey also explored adoption of NoSQL databases. As a newer emerging technology and one that is focused more on larger, distributed applications, it is unsurprising that NoSQL has not penetrated the enterprise to the same degree as cloud-based applications have. Less than 40% of the survey respondents currently support NoSQL implementations. And, NoSQL is clearly still an emerging technology. Of those running NoSQL applications, around 60% have been running for 2 years or less.

Similar to the reasons respondents offered for moving to the cloud, scalability is the primary driver to implement NoSQL.  Other attractive benefits are improved flexibility and decreased costs.

Robust cloud computing options have provided the possibility to move database-centric applications to off-premises locations. In addition, internet-scale applications such as social media, the networking of smartphones and tablets, and the growth of the Internet of Things have led to workloads that present challenges for relational databases.