Cross-Checking Multiple Data Sources Using Multiway Join in MapReduce

Table 1

Fat relations processing times (in milliseconds) on 8 compute nodes.

		Join 1: 2 out of 3 overlap			Join 3: 3 out of 4 overlap			Join 4: 3 out of 4; overlap
		# of reducers
		27	64	125	9	16	25	81
# of records	24,000	490840	351080	446020	38230	44970	61800	51780
	28,000	633200	466330	470890	61010	61200	67680	56420
	32,000	1068060	767200	694910	99940	95430	95280	60810
	36,000	1687300	1174030	1026560	159570	145420	145860	66490
	40,000	2620760	1852560	1484280	268030	245040	291930	74070