With both cloud computing and Big Data initiatives taking off in 2012, it is reasonable to expect the same for NoSQL in the next year.
Robert Greene, Vice-president for Technology at Versant gives his outlook for NoSQL in 2012:
”The relational database did an excellent job of managing moderately complex information models. The challenge is that as data grew, as it continues to do, this traditional database technology did not perform well and, on top of that, useful information models are getting even more complex. In 2010-2011 NoSQL ushered in a shift in architecture to deal with the performance-at-scale problem of the relational database, but unfortunately the world cannot ignore modern information model complexity and the early NoSQL technologies are scarcely addressing this point. In 2012 it will become clear that what is needed in NoSQL is a solution that addresses the scale issue, but also deals with the reality of complex, structured data. Ultimately, that is what we are producing from all of the ‘unstructured’ data we are capturing – an extracted layering of structure – so that the unstructured data becomes useful in solving real world analytical solutions.”
Given Mr. Greene’s focus on the technological challenges faced by NoSQL in 2012, especially with combining the scale of Big Data and the model complexity typical of sophisticated data analysis, it stands to reason that the successful business is one able to find the correct application for this still emerging database technology.
Conor O’Mahony, the Program Director for Database Software at IBM, follows up on this point:
“NoSQL shines in situations where incumbent technologies do not. For NoSQL to succeed in 2012, its proponents must convince IT organizations to select the best available technology for each project. One encouraging sign is that, as more organizations deploy NoSQL solutions, skills and knowledge are becoming more commonplace, which is in turn raising the IT industry’s comfort level with NoSQL technologies. Another encouraging sign is the general shift in the business world towards ‘use cases’ where NoSQL solutions shine, like applications for online business and social marketing.”
NoSQL facilitates systems with the scale of Big Data, but can it ever become well-suited for a wide array of BI applications? As it matures as a tool for businesses, what are some of the NoSQL trends to follow and challenges to overcome in 2012?
Hadoop Gets Some Competition
Nati Shalom, founder of cloud-based application virtualization provider, GigaSpaces, predicts that Hadoop’s domination of the Big Data space gets competition from NoSQL-based solutions like Cassandra and Mongo. Shalom feels Hadoop’s implementation complexity combined with its growing vendor fragmentation creates an opportunity for upstarts centered on NoSQL.
Security Issues Hamper NoSQL
Data security specialist, Imperva, published a list of nine cyber security trends for 2012, with NoSQL unfortunately making that list. Imperva feels the relatively new Big Data systems using NoSQL still lack the necessary security frameworks. Because of this weakness in security, Imperva feels enterprises will think twice before implementing fully integrated NoSQL-based solutions in 2012.
Oracle Gets in the NoSQL Game
In a move arguably legitimizing the entire NoSQL industry, longtime relational database vendor, Oracle, is betting on both Big Data and NoSQL in 2012. The company recently made its NoSQL-based database engine available for download.
The first quarter of 2012 sees the release of Oracle’s new “Big Data Appliance” that comes embedded with their NoSQL engine. Oracle’s NoSQL engine is based on BerkeleyDB and the company feels its implementation offers better ease-of-use than the competition.
In eWeek.com, Oracle’s Senior Vice-President, Andrew Mendelsohn, commented on their new Big Data and NoSQL initiatives:
“As customers look to manage the huge explosion in data from new and evolving sources, such as the Web, sensors, social networks and mobile applications, Oracle is helping them unlock the value of this data by providing a highly available, reliable and scalable NoSQL database environment.”
Microsoft Not Wanting to be Left Out
Displeased with seeing Oracle join the NoSQL party, Microsoft introduced their own Big Data and NoSQL initiatives in October, with the announcement of a Windows version of Hadoop. This further illustrates Nati Shalom’s earlier point about increased vendor fragmentation for the open source Big Data framework.
Redmond’s Hadoop implementation includes support for Windows Server and Azure as well as connectors to its flagship database product, SQL Server. Microsoft expects to preview the Azure functionality by the end of the 2011, with the Windows Server version to appear sometime in 2012. The SQL Server connectors are currently available.
Microsoft plans on submitting the code for its version of Hadoop to be considered for possible inclusion in the core open source project. Its recent partnership with Yahoo! spin-off, Hortonworks, is expected to facilitate the relationship with the core Hadoop team, as many of Hortonworks’ developers worked on that original project.
More In-Memory Data Grid and NoSQL Integration
In-memory data stores are a natural fit for the real-time analysis of Big Data, proven by successful implementations from Facebook and Twitter. GigaSpaces’ Nati Shalom feels that NoSQL is just right for this kind of application and expects to see more integration between the two in 2012.
The new company, Couchbase, created by the merger of CouchOne and Membase, is another example of the new synergies between in-memory data grids and NoSQL technology.
Hadoop Goes Enterprise
In an article for The Data Warehousing Institute, IBM’s Conor O’Mahony predicts big things for Apache Hadoop in 2012, namely its increased presence in the enterprise. O’Mahony remarks:
“Although Web-based businesses are quick to adopt Hadoop, general enterprise penetration has been modest to date, with a relatively limited number of early adopters leading the charge. However, there is no shortage of interest in Hadoop from organizations that see it as a way to cost-effectively tackle problems that had been difficult or impossible to solve. For instance, Hadoop offers the promise of analyzing the wealth of information in semi-structured or unstructured formats. I expect 2012 to be the year when Hadoop penetrates the enterprise in a meaningful way, and for the industry to converge on an early set of concrete use cases for Hadoop in the enterprise.”
O’Mahony’s expectation of further enterprise penetration for Hadoop, as supposed to its earlier use primarily in large scale web-based applications, exemplifies the increased relevance of Big Data and NoSQL. This does not necessarily contradict Nati Shalom’s prediction for increased competition for Hadoop; as both expectations support the growing importance of NoSQL.
NoSQL Consolidates its Dominance in the Cloud
Given the importance of rapid scalability in cloud computing applications, relational databases with their added complexity just do not provide the same performance as their NoSQL brethren.
Cloud computing expert, Mario Meir-Huber, wrote an article for SOA World magazine contrasting the two database approaches. NoSQL wins out for the responsiveness and scalability typical of cloud-based applications. Meir-Huber comments:
“For data warehousing and systems that require business intelligence to be in the database, SQL databases might be your choice. If you need high-responsive, scalable and flexible databases, NoSQL might be better for you.”
It is reasonable to assume as cloud computing gains relevance in 2012, so will NoSQL, as non-relational databases perform better than SQL in most cloud-based applications.
Despite some worries concerning security (normal for a still-emerging technology) there is little doubt 2012 will be a banner year for NoSQL. The increased dependence on scalable and flexible cloud-based solutions relegates relational databases to BI and legacy applications, making NoSQL preeminent in the cloud and for the future.