4 June 2015 Volha attended the ‘Big Data Everywhere’ event in London. The event was co-presented by MapR Technologies and Cisco, sponsored by Teradata, Protegrity and BMC Software and hosted by Hive London Meetup. The event was one in a series of international events and was largely introductory in nature. It was “aimed at helping organisations understand the technical, business and practical aspects of Hadoop and related technologies”. For information, Hadoop is a large-scale distributed computing technology that allows storing and handling large amounts of data “at a reasonable cost in a reasonable time” (Dunning & Friedman, 2015:6). It was originally “developed as an open source Apache Foundation project based on Google’s MapReduce paradigm” (Dunning & Friedman, 2015:7). Today it is widely used to enable big data-based projects in a variety of areas.
The programme was structured in the form of talks and a panel, which proceeded from more introductory ones on big data through to more applied ones on Hadoop and then to case studies. Talks in the first part of the programme focused on big data and data economy and reinforced the importance of data as a key asset for digital enterprises (“digital enterprises devour data” (Wentworth, MDW Advisors)) and of big data analytics as a way of securing a competitive advantage. They also emphasised the speed of change that characterises the data economy, with adoption rates of technology shortening dramatically and average age of a company in S&P 500 being just 10 years (McNeill, Google). Speed is also essential in terms of drawing insight from data analysis, because, “if you lose velocity, you lose the market” (McNeill, Google).
At the same time, according to Martin Willcox (Teradata), the novelty of the big data landscape has to do not so much with the increasing data volumes – after all, we have been living with this problem for a quite a while, – but with a newly acquired ability to measure behaviour directly instead of inferring it. To succeed, however, businesses need to be aware of a number of challenges, from multistructured datasets to the need for non-relational analytics for rapid iteration and discovery. This suggests that the future of digital enterprise is plural, based on multiple platforms and multiple information management strategies (Willcox, Teradata).
Continuing with the focus on strategies, suggested key goals for businesses included: leveraging all existing hardware; creating isolation of data sets in a different way; addressing the need to move code from one environment to another; and supporting real-time business continuity (Scott, MapR). As a result, “the next ‘last’ enterprise architecture” should be comprised of: dynamic computer resources; common storage platform; support for real-time applications; flexible programming models; deployment management; solution-based approach; and applications to operate business, all of which will eventually “become the traditional way of thinking” (Scott, MapR).
Finally, the case studies of businesses that had successfully implemented big data Apache Hadoop-based projects included a media company and a budget travel website, along with further examples from the financial sector and manufacturing.
Full agenda of the event is available at:
Dunning, T. and E. Friedman. 2015. Real-World Hadoop. Cambridge: O’Reilly.