Precognox Switched to Hadoop Based ETL Framework for Job Search Site Development
Precognox switched to a new, Hadoop based data collector framework at one of Hungary's premier job search website.
Kaposvar, Hungary (PRWEB) July 23, 2013
Precognox announced switching to a new, Hadoop based data collector framework at the Jobmonitor.hu job search website. The system is already live and working, serving one of Hungary’s most relevant sites of its kind.
According to Precognox, this solution enables easy adjustment to constantly growing demand at Jobmonitor with a relative decrease in costs and improvement in scalability.
The advantages of Hadoop-based distributed computing:
Hadoop is an open-source software architecture, developed by Apache, which supports data-intensive distributed applications.
Precognox uses the solution – amongst many – for a relative downscale of their resources at Jobmonitor in the future, as the new framework is flexible and easily scalable horizontally. It also minimizes data loss, since information in the Hadoop Distributed File System (HDFS) is redundantly shared in chunks among the members of the computer clusters involved. This, however, does not mean a decrease in performance as the MapReduce technology enables parallel processing of data on each individual computer.
The distributed ETL solution of Precognox uses HBase NoSQL database, supporting storing of a large amount of rows and columns.
Endre Jofoldi, CEO of Precognox: “Greg Rahn of Cloudera tweeted earlier this year: ‘So what do you do with Hadoop?’ ‘We do what everyone does with Hadoop… we talk about it’ – We are proud to say we managed to take the matter a step further and actually started using it.”
Jobmonitor is one of Hungary’s prime job search websites, developed by Precognox and owned by Sanoma Media – the country’s largest media distributor. Currently, up to 100 thousand queries are drawn each day in an increasing tendency but a vastly growing traffic is expected.
Precognox is a researcher and developer of intelligent, linguistic-based search, text mining and big data solutions. With a lengthy list of domestic and international clients – such as Gwabbit and the National Institutes of Health in the US or the Akademiai Kiado (Hungary's leading academic press company), etc. – it has a proven record for efficiency and reliability in custom search engine development and text mining. Senior members of the company bring in decades worth of experience in searching.
For the original version on PRWeb visit: http://www.prweb.com/releases/2013/7/prweb10941023.htm