Database Reference
In-Depth Information
Whether future inventions will be as groundbreaking as smartphone-based Star
Trek universal translators or as questionable as Internet-enabled toasters, the common
enabler for this inevitable, ubiquitous computing future is data. More specifically, this
vision requires that the huge amounts of data can be collected, stored, processed, and
analyzed in a useful way.
As a result, anyone who is involved with any aspect of data-processing technology
currently finds themselves in the middle of a very exciting but confusing era. To grok
Big Data means to have an awareness of the upcoming potential of the data generated
in the era of utility computing. Web-scale technology companies, including Yahoo!,
Google, Amazon, Facebook, and many others, have driven early innovations in dis-
tributed data processing, such as the MapReduce framework. Web companies were
forced to innovate in order to be successful. However, these use cases are just a precur-
sor to the potential data tidal wave that is coming soon.
Future predictions of technology are always doomed to face the judgment of retro-
spect; there is a disproven urban legend that IBM CEO Thomas J. Watson once said,
“I think there is a world market for about five computers.” Current trends in data are
driven by an inevitable need to provide easy easy-to-use tools to deal with growing
amounts of data being generated by both people and machines connected to the ever-
growing network.
“Big Data is Falling into the Trough of Disillusionment” reads the title of a blog post
by Gartner Research Director Svetlana Sicular.
1
In her post, Sicular claims to use a
methodology called the “Gartner Hype Cycle curve” to postulate that, as of January
2013, comments from a recent Hadoop conference indicate growing disdain for the
promise of Big Data.
Perhaps the most interesting thing about this blog post was neither the sensationalist
title nor the amusement that comes from pondering the veracity of the “Gartner Hype
Cycle.” Sicular brings up examples that solely revolve around Hadoop. The Apache
Hadoop project has become synonymous in the media with Big Data and for good
reason. The Hadoop ecosystem is huge, and there are many well-funded companies
who support its tools, including Cloudera, Hortonworks, MapR, and others. Estab-
lished database giants such as IBM and Oracle are mentioning Hadoop more and more.
Countless companies are working to improve, extend, and profit from Hadoop's reach.
Despite both the hype and success, Hadoop is not the be-all and end-all of large-
scale data-processing technologies. The sentiment expressed in Sicular's blog post points
toward some of the deficiencies that Apache Hadoop has in covering all aspects of data
collection, storage, and processing. Hadoop provides the ability to distribute storage
Search WWH ::
Custom Search