Network Data and Optimization
Mazin Gilbert, Mark A. Austin · 2017
138 Collecting and analyzing network event data can be a daunting processing challenge for today&s;s larger networks. For example, AT&T&s;s network in 2016 passes more than 118 petabytes (PB) per day; if each byte was to be treated as an “event,” it would imply a need to process, analyze, and act on approximately 1.4 trillion events per second. This sort of scale implies a need for a big data framework for event processing. This chapter describes some of the concepts for such data collection and analytics.