Nokia siemens networks

Very large scale data warehousing


NSN hardware underlies most of the networks run by carriers such as China Unicom and China Mobile, as well as the backbone of China's internet. In NSN's operating environment, 2.6 terabytes of data are generated per day.

Prior to our solution, this was simply stored in servers as no analysis could be run on. 

Using Cloudera's Hadoop solution, this data could be stored in its native text format and distributed over a cluster of over 100 blade servers. However, this still did not allow for quick analysis as queries would still take a day. Adding Infobright, a column store database allowed analytical queries to run in just seconds. 

Loading data from Hadoop to Infobright is seamless with our integrated algorithms. Users simply select what they need in Infobright, and the corresponding data is pulled from Hadoop and loaded into Infobright using batch loader tools in less than 10 minutes.