English subtitles

← Hashtables - Intro to Hadoop and MapReduce

Get Embed Code
2 Languages

Showing Revision 3 created 05/25/2016 by Udacity Robot.

  1. Historically, we'd probably use an associative array or a hash table
  2. to solve this problem in a traditional computing environment. The location
  3. would be the key and the sales for that store, the
  4. value. Then we'd process the input file one line at a
  5. time, adding each store as a key. What problems do you
  6. see with this approach? Say if you were running it on
  7. one terabyte of data. Do you think that it flat out
  8. won't work? Or could you run out of memory having to store
  9. this hash table. Could it take an excessively long
  10. time? Or will we not get the right answer?