April 20, 2012
The Big Data Drinking Game

Having watched almost 20 hours of STRATA and related talks, I propose the following drinking game for STRATA attendees:

  1. Every time Hadoop is mentioned, take a sip (trust me, you’ll go through a bottle in no time)
  2. Every time the words “web-scale”, “tera-scale” or, really, <anything>-scale are mentioned, drink a shot
  3. For every 1000 cores involved in computation, drink a shot
  4. For every terabyte per hour, drink a shot
  5. For every exabyte, finish the bottle.
  6. Drink a number of shots proportional to the square of the number of big-data technologies employed in a single system. For example: “We run Mahout over data stored in HBase running on a Hadoop cluster, and indexed in SOLR, while computing statistics using Hive”
    • Techologies employed = 5
    • Shots to be drunk = 25
    While this seems excessives, the resulting hangover will truly represent the headaches experienced by the speaker’s tech team 
  7. If the speaker builds custom, one-off hardware to run big data jobs, offer THEM the bottle. They’ll need it. 
  8. If the speaker has designed a custom programming language to do this work, buy them a single-malt in admiration of their bravery. If the language is based on LISP or SmallTalk, ask them who their drug dealer is and whether you can get some of whatever they’re smoking.

  1. skippy45s reblogged this from socialbrain
  2. socialbrain posted this