Improving I/O Efficiency in Hadoop-Based Massive Data Analysis Programs

Table 1

Statistics of the selected dataset from TPC-H [24].

Table name	Data size	The # of rows in a table
Customer	S ∗ 24 MB	S ∗ 150,000
Orders	S ∗ 171 MB	S ∗ 1,500,000
Lineitem	S ∗ 759 MB	S ∗ 6,001,215

S: scale factor.