Share, , Google Plus, Pinterest,

Print

Posted in:

Streamlining Hadoop Data for handy visualization

In the last data revolution of data-warehouses, organizations captured humongous amounts of data and are now struggling to handle it or find value out of it. Hadoop has been a big time enabler in letting these organizations bring value to this data. But, Hadoop is still quite complicated and its results have significant imperceptibility due to their crude form which are not friendly to the eye. This entails a large number of layers on top of Hadoop to make the data in desirable form and then visualize it, leading to design complications, dependency on expensive resources and restrictions in scalability and readiness. If Hadoop is not used, good and striking visualization of data can be accomplished but on a limited data-set, restricted by many factors like the local system’s memory and processing capabilities. As of now, people have to trade-off between the visualization quality, data size or speed.

There have been variety of efforts done towards solving this problem. Once of it is by Arcadia Data that provides native visualization capability on Hadoop data, providing functionalities like drag and drop of files, interactive data selection, etc. This will enable users to have the best of all worlds – better visualization, scalability and speed. Tableau connects directly to Hive, enabling users to extract data directly, though not natively, from Hadoop eco-system and further use it for analysis and visualization.

Platfora, rather, has created one comprehensive layer over Hadoop to enable end-to-end data processing capabilities for analytics. Atscale also creates readiness out of Hadoop data by making it available in format directly consumable by any BI tool of choice. It optimizes query in real time, enabling a quick execution over complete set of data. Jethrodata introduces a complete indexing methodology wherein select data is fully indexed so as to give superior execution time for queries on complete data-sets.

These accomplishments will, hopefully, bring big data closer to the wider base of its intended audience. Which solution shall be adopted in a specific case will be another discussion and will hover around the lower level details of the problem. But whichever suits most, it will mitigate the challenge of ‘change’ which today’s businesses face continually. Besides an ease, these developments will further call for a potential new range of products which would give out-of-the-box and domain specific functionalities harnessing Hadoop eco-systems as their information base. Such products can specifically focus on generic business challenges that need comprehensive and large data sets for solutions.

6 Comments

Leave a Reply
  1. With the emerging problem of Big Data analysis, Hadoop came as a fortune which gave a meaning to this huge amount of data. But with fortune comes some problems too which was also a part of Hadoop Eco-system. There were problems with hadoop too. But as they say, with technology comes updates for it’s betterment.
    Virtualization of hadoop came as a solution to this problem. Helping organisations to use hadoop in an effective and efficient manner. It has helped them in extracting the data directly from the hadoop Eco-system and use it for analysis.
    Companies are now finding it easy to analyse the data and are getting to know the advantages of hadoop in the industry.

  2. Visualization technologies for the hadoop based solutions is a big and revolutionary step which will definitely increase the audience of Hadoop. It will be good to those who have problems in understanding solutions from Hadoop Technologies. So it is a really nice article which tells us that work is being done in the field of visualization of Hadoop data.

  3. Its good article and also showing under the blanket problem with Big data technologies .with rise of Big data new-new framework and platform build over hadoop system also emerged to address specific problem associated with big data.And due to domain specific functionality of these technology in changing world is bigger issues for organization as its very expensive and time taking process to shift to another technology.their must be more development in direction of generic solutions.

  4. Thank you anirudh for sharing this article with us so that we could know about how to bring the big data too closer to their respective sources.

  5. Hopefully those visualization packages are user-friendly enough for those who have only conceptual level of, or even no knowledge of hadoop.

  6. Not sure about how resource heavy some of these solutions are but they definitely will alleviate lot of pain points for analysts with not so strong programming background. Thanks for sharing it.

Leave a Reply

Your email address will not be published. Required fields are marked *