Re:Ask HN: Cassandra, HBase, Hypertable, Voldemort. Lowest r...
by smallpotatoes on Tuesday June 29, 2010 @01:52PM
(#1472186)
An easy way to think about Zipfian distributions is something like "the most popular 30% of your data set is queried 70% of the time". Its basically the temporal locality property that a lot caching systems exploit.
For more look at http://en.wikipedia.org/wiki/Zipf%27s_law and www.icis.ntu.edu.sg/scs-ijit/1204/1204_6.pdf
My read patters will probably have locality but I have not yet figured out how to model the data in a key / value style to preserve the locality of the data. Can you describe what you mean by Zipfian distribution?