Skip to main content

Table 7 SSB execution times: bucketing by “Orderdate”, “Custkey”, “Suppkey” and “Partkey”

From: Evaluating partitioning and bucketing strategies for Hive-based Big Data Warehousing systems

 

SF = 30

SF = 100

SF = 300

SF = 30

SF = 100

SF = 300

HIVE

PRESTO

SS

SS-B

SS

SS-B

SS

SS-B

SS

SS-B

SS

SS-B

SS

SS-B

Q1.1

25

23

31

29

44

45

5

5

13

14

36

35

Q1.2

24

24

29

30

42

45

5

6

13

12

34

34

Q1.3

24

24

29

30

43

44

4

5

13

12

35

36

Q2.1

32

33

47

59

531

702

8

11

19

27

59

82

Q2.2

31

31

46

51

531

681

7

9

18

23

51

67

Q2.3

30

31

44

54

531

699

7

9

17

22

49

62

Q3.1

35

34

59

64

651

684

8

11

29

30

81

88

Q3.2

30

30

45

46

677

688

6

7

17

20

51

57

Q3.3

33

33

219

224

665

702

5

7

15

17

43

52

Q3.4

34

32

222

225

675

870

6

7

15

16

43

52

Q4.1

38

39

86

100

226

256

13

18

43

49

119

142

Q4.2

49

50

70

70

141

155

9

14

26

33

69

90

Q4.3

34

37

54

65

116

141

8

12

23

29

63

77

Total

420

420

982

1047

4874

5712

92

121

262

305

733

876

Diff.

 

0%

 

7%

 

17%

 

32%

 

16%

 

19%

  1. Italic values indicate the fastest processing time by query, workload, tool and data model