15
During the test, I have processed census data from CA and NY as described in Run HADOOP and JAQL on US Census 2000 Data using Hadoop 0.20.1 and Pig 0.5 on my Amazon EC2 51 nodes cluster!
Next I will start 100 spot instances (that is the default limit on the number of instances per an account) and test process full set of Census 2000.
