Slide 1

Slide 1 text

Joseph Blomstedt (@jtuple) Basho Technologies Seattle Scalability Meetup August 2013

Slide 2

Slide 2 text

2 Joe Blomstedt

Slide 3

Slide 3 text

3 Senior Engineer

Slide 4

Slide 4 text

Slide 5

Slide 5 text

5 @jtuple

Slide 6

Slide 6 text

6

Slide 7

Slide 7 text

7

Slide 8

Slide 8 text

8 Open Source

Slide 9

Slide 9 text

9 Open Source Apache 2 github.com/basho

Slide 10

Slide 10 text

10

Slide 11

Slide 11 text

11 Multi-DC Replication Support

Slide 12

Slide 12 text

Riak Basics 12

Slide 13

Slide 13 text

13

Slide 14

Slide 14 text

14 Key / Value Secondary Indexes Full Text Search Map/Reduce Counters

Slide 15

Slide 15 text

15 Key / Value Secondary Indexes Full Text Search Map/Reduce Counters

Slide 16

Slide 16 text

16 PUT GET DELETE

Slide 17

Slide 17 text

17 users/joe PUT

Slide 18

Slide 18 text

18 users/joe GET

Slide 19

Slide 19 text

19 users/joe DELETE

Slide 20

Slide 20 text

20 Key / Value Secondary Indexes Full Text Search Map/Reduce Counters

Slide 21

Slide 21 text

21 user/1234 PUT {first: “John”, last: “Doe”, age: 40}

Slide 22

Slide 22 text

22 user/1234 PUT {first: “John”, last: “Doe”, age: 40} first_bin: “John” last_bin: “Doe” age_int: 40

Slide 23

Slide 23 text

23 pics/pic42 PUT animal_bin: cat cuteness_int: 10

Slide 24

Slide 24 text

24 Tagged Fields Exact/Range Single Index Streaming Pagination

Slide 25

Slide 25 text

25 Key / Value Secondary Indexes Full Text Search Map/Reduce Counters

Slide 26

Slide 26 text

26 Automatic Indexing JSON XML Text Custom

Slide 27

Slide 27 text

27 Querying Exact Match Wildcard Range

Slide 28

Slide 28 text

28 Querying Grouping Prefix matching Term Boosting

Slide 29

Slide 29 text

29 Querying AND OR NOT

Slide 30

Slide 30 text

30 Key / Value Secondary Indexes Full Text Search Map/Reduce Counters

Slide 31

Slide 31 text

31 Use Cases Filtering Aggregation Extraction

Slide 32

Slide 32 text

32 M/R Input Known Keys 2nd Index Query Full Text Query

Slide 33

Slide 33 text

33 Distributed Mappers: data nodes Reducers: coordinator

Slide 34

Slide 34 text

34 Key / Value Secondary Indexes Full Text Search Map/Reduce Counters

Slide 35

Slide 35 text

35 Why do people use Riak?

Slide 36

Slide 36 text

Big Data 36

Slide 37

Slide 37 text

Big Data 37 Scalable Distributed

Slide 38

Slide 38 text

Big Data 38 5 nodes 100 nodes

Slide 39

Slide 39 text

Critical Data 39

Slide 40

Slide 40 text

Critical Data 40 Big Medium small

Slide 41

Slide 41 text

Fault Tolerance 41

Slide 42

Slide 42 text

Availability 42

Slide 43

Slide 43 text

CAP Theorem 43

Slide 44

Slide 44 text

44 Partition-tolerance Consistency Availability

Slide 45

Slide 45 text

45 Partition-tolerance Consistency Availability

Slide 46

Slide 46 text

46 Partition-tolerance Consistency Availability CP AP

Slide 47

Slide 47 text

47 Partition-tolerance Consistency Availability CP AP

Slide 48

Slide 48 text

48 C/P Strict Quorum A/P Sloppy Quorum A/P

Slide 49

Slide 49 text

49 C/P Strict Quorum A/P Sloppy Quorum A/P

Slide 50

Slide 50 text

50 Node 1 Node 2 Node 3 Node 4 Node 5 client client client

Slide 51

Slide 51 text

51 Node 1 Node 2 Node 3 Node 4 Node 5 client client client

Slide 52

Slide 52 text

52 Node 1 Node 2 Node 3 Node 4 Node 5 client client client

Slide 53

Slide 53 text

53 Node 1 Node 2 Node 3 Node 4 Node 5 client client client

Slide 54

Slide 54 text

54 C/P Strict Quorum A/P Sloppy Quorum A/P

Slide 55

Slide 55 text

55 Node 1 Node 2 Node 3 Node 4 Node 5 client client client

Slide 56

Slide 56 text

56 Node 1 Node 2 Node 3 Node 4 Node 5 client client client

Slide 57

Slide 57 text

57 Node 1 Node 2 Node 3 Node 4 Node 5 client client client

Slide 58

Slide 58 text

58 Node 1 Node 2 Node 3 Node 4 Node 5 client client client

Slide 59

Slide 59 text

59 C/P Strict Quorum A/P Sloppy Quorum A/P

Slide 60

Slide 60 text

60 Node 1 Node 2 Node 3 Node 4 Node 5 client client client

Slide 61

Slide 61 text

61 Node 1 Node 2 Node 3 Node 4 Node 5 client client client

Slide 62

Slide 62 text

62 Node 1 Node 2 Node 3 Node 4 Node 5 client client client

Slide 63

Slide 63 text

63 Node 1 Node 2 Node 3 Node 4 Node 5 client client client

Slide 64

Slide 64 text

64 Node 1 Node 2 Node 3 Node 4 Node 5 client client client client client

Slide 65

Slide 65 text

Eventual Consistency 65

Slide 66

Slide 66 text

66 A A A

Slide 67

Slide 67 text

67 A A A

Slide 68

Slide 68 text

68 A A A B

Slide 69

Slide 69 text

69 A A A B

Slide 70

Slide 70 text

70 A A A B B B B

Slide 71

Slide 71 text

71 A A A

Slide 72

Slide 72 text

72 A A A B C

Slide 73

Slide 73 text

73 A A A B C

Slide 74

Slide 74 text

74 A A A B {B,C} {B,C} {B,C} C

Slide 75

Slide 75 text

75 Write Once Immutable Last Write Wins Business Rules CRDTs/Monotonicity

Slide 76

Slide 76 text

76 Write Once Immutable Last Write Wins Business Rules CRDTs/Monotonicity

Slide 77

Slide 77 text

77 Write Once Immutable Last Write Wins Business Rules CRDTs/Monotonicity

Slide 78

Slide 78 text

78 Write Once Immutable Last Write Wins Business Rules CRDTs/Monotonicity

Slide 79

Slide 79 text

79 Write Once Immutable Last Write Wins Business Rules CRDTs/Monotonicity

Slide 80

Slide 80 text

80 Write Once Immutable Last Write Wins Business Rules CRDTs/Monotonicity

Slide 81

Slide 81 text

81 Write Once Immutable Last Write Wins Business Rules “Set Union”

Slide 82

Slide 82 text

82 {1} {1} {1}

Slide 83

Slide 83 text

83 {1} {1} {1}

Slide 84

Slide 84 text

84 {1} {1} {1} {1,2} {1,3} {1,2,4}

Slide 85

Slide 85 text

85 {1} {1} {1} {1,2} {1,3} {1,2,4}

Slide 86

Slide 86 text

86 {1} {1} {1} {1,2} {1,2,3,4} {1,2,3,4} {1,2,3,4} {1,3} {1,2,4}

Slide 87

Slide 87 text

Clustering 87

Slide 88

Slide 88 text

88 Consistent Hashing Sloppy Quorums Hinted Handoff

Slide 89

Slide 89 text

89 0 2160

Slide 90

Slide 90 text

90 0 2160

Slide 91

Slide 91 text

91

Slide 92

Slide 92 text

92 Node 2 Node 1 Node 3

Slide 93

Slide 93 text

93 Node 2 Node 1 Node 3 Node 4

Slide 94

Slide 94 text

94 Node 2 Node 1 Node 3 Node 4 SHA1(Key)

Slide 95

Slide 95 text

95 Node 2 Node 1 Node 3 Node 4 SHA1(Key) primary replicas

Slide 96

Slide 96 text

96 Node 2 Node 1 Node 3 Node 4 primary + fallback replicas

Slide 97

Slide 97 text

97 Node 2 Node 1 Node 3 Node 4 handoff

Slide 98

Slide 98 text

Self Healing 98

Slide 99

Slide 99 text

99 Here Today != Here Tomorrow

Slide 100

Slide 100 text

100 Read Repair

Slide 101

Slide 101 text

101 Active Anti Entropy

Slide 102

Slide 102 text

102 Node 2 Node 1 Node 3 Node 4 AAE exchange

Slide 103

Slide 103 text

103 Active Anti Entropy Realtime Persistent Rebuilt Weekly

Slide 104

Slide 104 text

Upcoming 104

Slide 105

Slide 105 text

105 Data Types (CRDTs) Strong Consistency (C/P) Solr Integration

Slide 106

Slide 106 text

CRDTs 106

Slide 107

Slide 107 text

107

Slide 108

Slide 108 text

108

Slide 109

Slide 109 text

109 Counters Sets Maps

Slide 110

Slide 110 text

110 Counters Sets Maps

Slide 111

Slide 111 text

111 1 1 1

Slide 112

Slide 112 text

112 1 1 1

Slide 113

Slide 113 text

113 1 1 1 2 4 +1 +3

Slide 114

Slide 114 text

114 1 1 1 2 4 +1 +3

Slide 115

Slide 115 text

115 1 1 1 2 {2,4} {2,4} {2,4} 4 +1 +3

Slide 116

Slide 116 text

116 (a,1) (a,1) (a,1)

Slide 117

Slide 117 text

117 (a,1) (a,1) (a,1)

Slide 118

Slide 118 text

118 (a,1) +1 +3 (a,1) (a,1) (a,2) (a,1),(b,3)

Slide 119

Slide 119 text

119 (a,1) +1 +3 (a,1) (a,1) (a,2) (a,1),(b,3)

Slide 120

Slide 120 text

120 (a,1) +1 +3 (a,1) (a,1) (a,2) (a,1),(b,3) (a,2),(b,3) (a,2),(b,3) (a,1),(b,3)

Slide 121

Slide 121 text

121 (a,1) +1 +3 (a,1) (a,1) (a,2) (a,1),(b,3) (a,2),(b,3) (a,2),(b,3) (a,1),(b,3) =5

Slide 122

Slide 122 text

122 Counters Sets Maps

Slide 123

Slide 123 text

123 Counters Sets Maps

Slide 124

Slide 124 text

Consistency 124

Slide 125

Slide 125 text

125 A/P or C/P Per Bucket Choice

Slide 126

Slide 126 text

126 Single Key Optimistic CAS Counters/Sets/Maps

Slide 127

Slide 127 text

127 Multi Paxos Disjoint Ensembles Scalable

Slide 128

Slide 128 text

Node 2 Node 1 Node 3

Slide 129

Slide 129 text

129 1 2 3 4 5 6 7 123

Slide 130

Slide 130 text

130 1 2 3 4 5 6 7 123 234

Slide 131

Slide 131 text

131 1 2 3 4 5 6 7 123 234 345

Slide 132

Slide 132 text

132 1 2 3 4 5 6 7 123 234 345 456

Slide 133

Slide 133 text

133 1 2 3 4 5 6 7 123 234 345 456 567...

Slide 134

Slide 134 text

134 1 2 3 4 5 6 7 123 Ensembles 234 345 456 567...

Slide 135

Slide 135 text

135 Ensemble Indepedent Paxos Group Leader Election Serializability

Slide 136

Slide 136 text

136 write 2 1 1 1 write 2 leader

Slide 137

Slide 137 text

137 write 2 2 2 2 write 2 leader 2 X

Slide 138

Slide 138 text

138 3 3 3 write 3 leader 3

Slide 139

Slide 139 text

139 3 3 3 leader

Slide 140

Slide 140 text

140 +1 1 1 1 +1 leader

Slide 141

Slide 141 text

141 2 2 2 +1 leader 2

Slide 142

Slide 142 text

142 3 3 3 leader 3

Slide 143

Slide 143 text

143 3 3 3 leader

Slide 144

Slide 144 text

Riak + Solr 144

Slide 145

Slide 145 text

145 Next Generation Search Full Distributed Solr Embedded in Riak Active Anti Entropy

Slide 146

Slide 146 text

End 146

Slide 147

Slide 147 text

147 We hiring!

Slide 148

Slide 148 text

No content

Slide 149

Slide 149 text

No content

Slide 150

Slide 150 text

ricon.io

Slide 151

Slide 151 text

Questions? 151