Enterprises need to use more than Hadoop to analyse big data, says research

One in three enterprises are currently using Hadoop software for big data implementations, with another third saying they were planning to deploy within 12 months.

That’s the top-line finding from an IDC study, sponsored by Red Hat, which examines companies’ big data strategies – but found that many organisations are “forced” into augmenting Hadoop deployments with commercial add-ons because it’s not enterprise-class resilient yet.

The reasons for using Hadoop were interesting too; analysis of transactional data and online customer behaviour data were the most important use cases. More than half of respondents said that deploying Hadoop helped improve customer satisfaction.

Yet the overall consensus from the report was that Hadoop wasn’t enough on its own for big data analysis, leveraging external storage software to enhance the functionality of Hadoop. Not surprisingly, Red Hat got special mention as a storage vendor.

Companies were also using other solutions. Nearly 39% of respondents said they used NoSQL databases, such as HBase, Cassandra and MongoDB, whilst a further 36% said they used MPP databases, such as Greenplum and Vertica.

“Businesses will become data driven to the point where they adopt a hawkish, discovery-oriented mindset,” the report notes, going on to note the ‘profound’ effect this will have – in particular moving from a capex-heavy model to an opex-heavy one.

This latest research agrees with Ken Rudin, head of analytics at Facebook, who told the Strata + Hadoop World conference in New York that companies need to expand beyond Hadoop to make sense of unstructured big data.

As reported by Computer World, Rudin told attendees: “In reality, big data should include Hadoop and relational, and any other technology that is suitable for the task at hand.”

Big data company Splunk recently made its Hadoop-crunching tool Hunk generally available, and according to Matt Davies, Splunk EMEA product marketing director, Hadoop is a “massive deal” in the US but there remains “healthy scepticism” in EMEA.

It’s a difficult choice for companies looking to make the most of big data as to which database technology they use. Do they go NoSQL, do they go relational?

For Davies, there won’t be a clear winner just yet.

“I can’t imagine there’ll be a place where SQL databases and relational databases ever disappear,” he said. “But I think it’s interesting looking at some of the very modern application architectures and seeing how they aren’t using relational databases.”

The full study, entitled ‘Trends in Enterprise Hadoop Deployments’, can be found here. But what do you think of the research? in hearing industry leaders discuss subjects like this and sharing their use-cases? Attend the co-located IoT Tech Expo, Blockchain Expo, AI & Big Data Expo and Cyber Security & Cloud Expo World Series with upcoming events in Silicon Valley, London and Amsterdam and explore the future of enterprise technology.


Related Stories

Leave a comment


This will only be used to quickly provide signup information and will not allow us to post to your account or appear on your timeline.

6 Nov 2013, 2:25 p.m.

Interesting article - thanks! I've posted on some possible reasons why NoSQL is moving past RDBMS MPP/column DB's here: