Upgrade to Pro — share decks privately, control downloads, hide ads and more …

RICON West 2012: Bringing Consistency to Riak (...

Avatar for Joseph Blomstedt Joseph Blomstedt
October 30, 2013
Programming
3
780

RICON West 2012: Bringing Consistency to Riak (Part 2)

Avatar for Joseph Blomstedt

Joseph Blomstedt

October 30, 2013
Tweet

More Decks by Joseph Blomstedt

See All by Joseph Blomstedt
Data Structure Adventures for Fun, Profit, and Performance
Avatar for Joseph Blomstedt jtuple
2
350
Riak -- Seattle Scalability
Avatar for Joseph Blomstedt jtuple
2
470
Hansei: Property-based Development of Concurrent Systems
Avatar for Joseph Blomstedt jtuple
3
2.2k
Riak 1.2 Webinar Preview
Avatar for Joseph Blomstedt jtuple
1
280
Test First Construction of Distributed Systems
Avatar for Joseph Blomstedt jtuple
3
1.1k

Other Decks in Programming

See All in Programming
フロントエンド開発の勘所 -複数事業を経験して見えた判断軸の違い-
Avatar for Yoichiro Kubo heimusu
7
2.6k
Canon EOS R50 V と R5 Mark II 購入でみえてきた最近のデジイチ VR180 事情、そして VR180 静止画に活路を見出すまで
Avatar for kazuhiro hara karad
0
150
re:Invent 2025 のイケてるサービスを紹介する
Avatar for maroon1st maroon1st
0
170
Pythonではじめるオープンデータ分析〜書籍の紹介と書籍で紹介しきれなかった事例の紹介〜
Avatar for Ryo YOSHI welliving
3
800
CSC307 Lecture 03
Avatar for Javier Gonzalez-Sanchez javiergs
PRO
1
480
AgentCoreとHuman in the Loop
Avatar for Har1101 har1101
5
180
Basic Architectures
Avatar for Denys Poltorak denyspoltorak
0
370
AtCoder Conference 2025
Avatar for shindannin shindannin
0
960
CSC307 Lecture 01
Avatar for Javier Gonzalez-Sanchez javiergs
PRO
0
670
なるべく楽してバックエンドに型をつけたい！(楽とは言ってない)
Avatar for HIBIKI CUBE hibiki_cube
0
120
[AI Engineering Summit Tokyo 2025] LLMは計画業務のゲームチェンジャーか？ 最適化業務における活⽤の可能性と限界
Avatar for terry-u16 terryu16
2
360
dchart: charts from deck markup
Avatar for Anthony Starks ajstarks
3
960

Featured

See All Featured
[SF Ruby Conf 2025] Rails X
Avatar for Vladimir Dementyev palkan
0
710
Visualization
Avatar for Eitan Lees eitanlees
150
16k
Large-scale JavaScript Application Architecture
Avatar for Addy Osmani addyosmani
515
110k
[RailsConf 2023 Opening Keynote] The Magic of Rails
Avatar for Eileen M. Uchitelle eileencodes
31
9.8k
VelocityConf: Rendering Performance Case Studies
Avatar for Addy Osmani addyosmani
333
24k
Agile that works and the tools we love
Avatar for Rasmus Luckow-Nielsen rasmusluckow
331
21k
BBQ
Avatar for Matthew Crist matthewcrist
89
10k
Design and Strategy: How to Deal with People Who Don’t "Get" Design
Avatar for Morgane Peng morganepeng
133
19k
Building Flexible Design Systems
Avatar for Yesenia Perez-Cruz yeseniaperezcruz
330
40k
XXLCSS - How to scale CSS and keep your sanity
Avatar for Zaharenia Atzitzikaki sugarenia
249
1.3M
The agentic SEO stack - context over prompts
Avatar for schlessera schlessera
0
590
Discover your Explorer Soul
Avatar for Emna emna__ayadi
2
1k

Transcript

  1. Joseph Blomstedt (@jtuple) Basho Technologies Bringing Consistency To Riak (Part

    2) Tuesday, October 29, 13

  2. CAP Theorem 2 Tuesday, October 29, 13

  3. 3 Partition-tolerance Consistency Availability Tuesday, October 29, 13

  4. 4 Partition-tolerance Consistency Availability Tuesday, October 29, 13

  5. 5 Partition-tolerance Consistency Availability CP AP Tuesday, October 29, 13

  6. 6 Partition-tolerance Consistency Availability CP AP Tuesday, October 29, 13

  7. 7 Partition-tolerance Consistency Availability CP AP Tuesday, October 29, 13

  8. 8 C/P Strict Quorum A/P Sloppy Quorum A/P Tuesday, October

    29, 13

  9. 9 C/P Strict Quorum A/P Sloppy Quorum A/P Tuesday, October

    29, 13

  10. 10 Node 1 Node 2 Node 3 Node 4 Node

    5 client client client Tuesday, October 29, 13

  11. 11 Node 1 Node 2 Node 3 Node 4 Node

    5 client client client Tuesday, October 29, 13

  12. 12 Node 1 Node 2 Node 3 Node 4 Node

    5 client client client Tuesday, October 29, 13

  13. 13 Node 1 Node 2 Node 3 Node 4 Node

    5 client client client Tuesday, October 29, 13

  14. 14 C/P Strict Quorum A/P Sloppy Quorum A/P Tuesday, October

    29, 13

  15. 15 Node 1 Node 2 Node 3 Node 4 Node

    5 client client client Tuesday, October 29, 13

  16. 16 Node 1 Node 2 Node 3 Node 4 Node

    5 client client client Tuesday, October 29, 13

  17. 17 Node 1 Node 2 Node 3 Node 4 Node

    5 client client client Tuesday, October 29, 13

  18. 18 Node 1 Node 2 Node 3 Node 4 Node

    5 client client client Tuesday, October 29, 13

  19. 19 C/P Strict Quorum A/P Sloppy Quorum A/P Tuesday, October

    29, 13

  20. 20 Node 1 Node 2 Node 3 Node 4 Node

    5 client client client Tuesday, October 29, 13

  21. 21 Node 1 Node 2 Node 3 Node 4 Node

    5 client client client Tuesday, October 29, 13

  22. 22 Node 1 Node 2 Node 3 Node 4 Node

    5 client client client Tuesday, October 29, 13

  23. 23 Node 1 Node 2 Node 3 Node 4 Node

    5 client client client Tuesday, October 29, 13

  24. 24 Node 1 Node 2 Node 3 Node 4 Node

    5 client client client client client Tuesday, October 29, 13

  25. Eventual Consistency 25 Tuesday, October 29, 13

  26. 26 A A A Tuesday, October 29, 13

  27. 27 A A A Tuesday, October 29, 13

  28. 28 A A A B Tuesday, October 29, 13

  29. 29 A A A B Tuesday, October 29, 13

  30. 30 A A A B B B B Tuesday, October

    29, 13

  31. 31 A A A Tuesday, October 29, 13

  32. 32 A A A B C Tuesday, October 29, 13

  33. 33 A A A B C Tuesday, October 29, 13

  34. 34 A A A B {B,C} {B,C} {B,C} C Tuesday,

    October 29, 13

  35. 35 Write Once Immutable Last Write Wins Business Rules CRDTs/Monotonicity

    Tuesday, October 29, 13

  36. 36 Write Once Immutable Last Write Wins Business Rules CRDTs/Monotonicity

    Tuesday, October 29, 13

  37. 37 Write Once Immutable Last Write Wins Business Rules CRDTs/Monotonicity

    Tuesday, October 29, 13

  38. 38 Write Once Immutable Last Write Wins Business Rules CRDTs/Monotonicity

    Tuesday, October 29, 13

  39. 39 Write Once Immutable Last Write Wins Business Rules CRDTs/Monotonicity

    Tuesday, October 29, 13

  40. 40 Write Once Immutable Last Write Wins Business Rules CRDTs/Monotonicity

    Tuesday, October 29, 13

  41. Strong Consistency 41 Tuesday, October 29, 13

  42. Strong Consistency 42 Why? Tuesday, October 29, 13

  43. Strong Consistency 43 Recency Tuesday, October 29, 13

  44. Strong Consistency 44 Recency Partial Writes Tuesday, October 29, 13

  45. Strong Consistency 45 Recency Partial Writes Atomicity Tuesday, October 29,

    13

  46. 46 Recency Partial Writes Atomicity Tuesday, October 29, 13

  47. 47 Recency Partial Writes Atomicity Tuesday, October 29, 13

  48. 48 Eventual consistency is great Tuesday, October 29, 13

  49. 49 But, when is eventual? Tuesday, October 29, 13

  50. 50 Do I have the most recent value? Tuesday, October

    29, 13

  51. 51 CRDTs don’t help Tuesday, October 29, 13

  52. 52 (a,1) (a,1) (a,1) =1 Tuesday, October 29, 13

  53. 53 (a,1) (a,1) (a,1) Tuesday, October 29, 13

  54. 54 (a,1) +1 +3 (a,1) (a,1) (a,2) (a,1),(b,3) =2 =4

    Tuesday, October 29, 13

  55. 55 (a,1) +1 +3 (a,1) (a,1) (a,2) (a,1),(b,3) Tuesday, October

    29, 13

  56. 56 (a,1) +1 +3 (a,1) (a,1) (a,2) (a,1),(b,3) (a,2),(b,3) (a,2),(b,3)

    (a,2),(b,3) Tuesday, October 29, 13

  57. 57 (a,1) +1 +3 (a,1) (a,1) (a,2) (a,1),(b,3) (a,2),(b,3) (a,2),(b,3)

    (a,2),(b,3) =5 Tuesday, October 29, 13

  58. 58 (a,1) +1 +3 (a,1) (a,1) (a,2) (a,1),(b,3) =2 =4

    Tuesday, October 29, 13

  59. 59 Recency Partial Writes Atomicity Tuesday, October 29, 13

  60. 60 A write B (fail) A A B A A

    Tuesday, October 29, 13

  61. 61 B A A Tuesday, October 29, 13

  62. 62 B A A read A read A read A

    Tuesday, October 29, 13

  63. 63 B A A read A read A read A

    Tuesday, October 29, 13

  64. 64 B A A read A read A read A

    read B Tuesday, October 29, 13

  65. 65 Recency Partial Writes Atomicity Tuesday, October 29, 13

  66. Strong Consistency 66 Tuesday, October 29, 13

  67. Strong Consistency 67 What does mean for Riak 2.0? Tuesday,

    October 29, 13

  68. 68 Conditional single key atomic operations Tuesday, October 29, 13

  69. 69 No siblings Tuesday, October 29, 13

  70. 70 get sees most recent put Tuesday, October 29, 13

  71. 71 get/modify/put fails if object changed Tuesday, October 29, 13

  72. 72 get/modify/put fails if object changed (eg. concurrent put) Tuesday,

    October 29, 13

  73. 73 puts w/o vclock fails if object exists Tuesday, October

    29, 13

  74. 74 partial writes resolved on read Tuesday, October 29, 13

  75. 75 Consensus Tuesday, October 29, 13

  76. 76 Paxos Tuesday, October 29, 13

  77. 77   1RGH    1RGH  

     1RGH    1  SUHSDUH 1   SURPLVH 1 9 %  SURPLVH 1 9 &  9 1 I 9 $  9 %  9 &  FRPPLW 1 9 1    DFFHSW 1  Tuesday, October 29, 13

  78. 78 Rinse/repeat for each request Tuesday, October 29, 13

  79. 79 2 round trips/request Tuesday, October 29, 13

  80. 80 Multi-Paxos Tuesday, October 29, 13

  81. 81 First Request Tuesday, October 29, 13

  82. 82   1RGH    1RGH  

     1RGH    1 ,   SUHSDUH 1 ,   SURPLVH 1 , 9 %  SURPLVH 1 , 9 &  9 1 I 9 $  9 %  9 &  FRPPLW 1 , 9 1    DFFHSW 1 ,  Tuesday, October 29, 13

  83. 83 Each Additional Request Tuesday, October 29, 13

  84. 84   1RGH    1RGH  

     1RGH    ,  FRPPLW 1 , 9    DFFHSW 1 ,  Tuesday, October 29, 13

  85. 85 1 round trip/request (common case) Tuesday, October 29, 13

  86. 86 Problem Shipping entire state each request is expensive Tuesday,

    October 29, 13

  87. 87 Solution Paxos + Replicated Log Tuesday, October 29, 13

  88. 88 Problem Now I have N problems Tuesday, October 29,

    13

  89. 89 Log recovery Log trimming Rollup Snapshots Fault Recovery Tuesday,

    October 29, 13

  90. 90 Choose your own adventure... Tuesday, October 29, 13

  91. 91 Better Solution Build log replication into protocol Tuesday, October

    29, 13

  92. 92 Better Solution ZK Atomic Broadcast Raft Tuesday, October 29,

    13

  93. Zab 93 Tuesday, October 29, 13

  94. 94 Tuesday, October 29, 13

  95. 95 Tuesday, October 29, 13

  96. 96 Tuesday, October 29, 13

  97. 97 Tuesday, October 29, 13

  98. Raft 98 Tuesday, October 29, 13

  99. 99 Tuesday, October 29, 13

  100. 100 raftconsensus.github.io Tuesday, October 29, 13

  101. 101 Text Tuesday, October 29, 13

  102. Back to Riak 102 Tuesday, October 29, 13

  103. 103 Key/Value Keys are independent Active Anti-Entropy Tunable backends Tuesday,

    October 29, 13

  104. 104 Each key is independent state Tuesday, October 29, 13

  105. 105 Simple multi-paxos per key Tuesday, October 29, 13

  106. 106 1B keys = 1B consensus groups? Tuesday, October 29,

    13

  107. 107 No Tuesday, October 29, 13

  108. 108 Consensus group per preflist (replica set) Tuesday, October 29,

    13

  109. 109 Emulate paxos per key Tuesday, October 29, 13

  110. Node 0 Node 1 Node 2 Tuesday, October 29, 13

  111. 111 1 2 3 4 5 6 7 123 Tuesday,

    October 29, 13

  112. 112 1 2 3 4 5 6 7 123 234

    Tuesday, October 29, 13

  113. 113 1 2 3 4 5 6 7 123 234

    345 Tuesday, October 29, 13

  114. 114 1 2 3 4 5 6 7 123 234

    345 456 Tuesday, October 29, 13

  115. 115 1 2 3 4 5 6 7 123 234

    345 456 567 ... Tuesday, October 29, 13

  116. 116 1 2 3 4 5 6 7 123 234

    345 456 567 Ensembles ... Tuesday, October 29, 13

  117. 117 64 partition ring = 64 ensembles Tuesday, October 29,

    13

  118. 118 Each Ensemble Elects leader Establishes epoch Supports get/put ops

    Tuesday, October 29, 13

  119. 119 Establish a new epoch Tuesday, October 29, 13

  120. 120   1RGH    1RGH  

     1RGH    1 ,   SUHSDUH 1 ,   SURPLVH 1 , 9 %  SURPLVH 1 , 9 &  9 1 I 9 $  9 %  9 &  FRPPLW 1 , 9 1    DFFHSW 1 ,  Tuesday, October 29, 13

  121. 121 consensus state epoch sequence membership leader Tuesday, October 29,

    13

  122. 122 K/V objects epoch sequence key value Tuesday, October 29,

    13

  123. 123 GET leader reads local object if obj.epoch old: refresh

    reply w/ val Tuesday, October 29, 13

  124. 124   1RGH    1RGH  

     1RGH    REMHSRFK  HSRFK  JHW .H\   UHSO\ (SRFK %  6HT %  9DO %  UHSO\ (SRFK &  6HT &  9DO &  9DO ODWHVW 9DO $  9DO %  9DO & 9DOHSRFK HSRFK  ZULWH (SRFK 6HT 9DO    DFN (SRFK 6HT  Tuesday, October 29, 13

  125. 125   1RGH    1RGH  

     1RGH    REMHSRFK HSRFK  5HSO\ ORFDOBJHW .H\  Tuesday, October 29, 13

  126. 126 2 roundtrips/get (worst) 0 roundtrip/get (best) Tuesday, October 29,

    13

  127. 127 PUT leader reads local object if obj.epoch old: refresh

    if modify(obj) false: fail commit modified obj reply ok Tuesday, October 29, 13

  128. 128   1RGH    1RGH  

     1RGH    REMHSRFK  HSRFK  JHW .H\   UHSO\ (SRFK %  6HT %  9DO %  UHSO\ (SRFK &  6HT &  9DO &  /DWHVW ODWHVW 9DO $  9DO %  9DO & 9DO PRGLI\ /DWHVW  ZULWH (SRFK 6HT 9DO    DFN (SRFK 6HT  Tuesday, October 29, 13

  129. 129   1RGH    1RGH  

     1RGH    REMHSRFK HSRFK  /DWHVW ORFDOBJHW .H\ 9DO PRGLI\ /DWHVW  ZULWH (SRFK 6HT 9DO    DFN (SRFK 6HT  Tuesday, October 29, 13

  130. 130 2 roundtrips/put (worst) 1 roundtrip/put (best) Tuesday, October 29,

    13

  131. 131 Leader abandons leadership if any quorum operation ever fails

    Tuesday, October 29, 13

  132. 132 Which forces new epoch to be established Tuesday, October

    29, 13

  133. 133 Partial Writes Tuesday, October 29, 13

  134. failed partial write X (2) X (2) X (2) X

    (2) X (2) Y (2) epoch 2 epoch 3 Tuesday, October 29, 13

  135. read / rewrite / reply X X (2) X (2)

    Y (2) X (3) X (3) Y (2) epoch 3 epoch 3 Tuesday, October 29, 13

  136. X (3) X (3) Y (2) X (3) X (3)

    X (3) read / repair / reply X epoch 3 epoch 3 Tuesday, October 29, 13

  137. Usage 137 Tuesday, October 29, 13

  138. 138 AP or CP per bucket type Tuesday, October 29,

    13

  139. 139 consistent = true Tuesday, October 29, 13

  140. 140 $ riak-admin bucket-type create strong \ '{"props": {"consistent": true}}'

    strong created Tuesday, October 29, 13

  141. 141 $ riak-admin bucket-type activate strong strong has been activated

    Tuesday, October 29, 13

  142. 142 > riakc_pb_socket:get(Socket, {<<"strong">>, <<"bucket">>}, <<"key">>). {error,notfound} Tuesday, October 29,

    13

  143. 143 > Obj = riakc_obj:new({<<"strong">>, <<"bucket">>}, <<"key">>, <<"1">>)). > riakc_pb_socket:put(Socket,

    Obj). ok Tuesday, October 29, 13

  144. 144 > Obj2 = riakc_obj:new({<<"strong">>, <<"bucket">>}, <<"key">>, <<"2">>)). > riakc_pb_socket:put(Socket,

    Obj2). {error, failed} Tuesday, October 29, 13

  145. 145 {ok, Obj3} = riakc_pb_socket:get(Socket, {<<"strong">>, <<"bucket">>}, <<"key">>). Tuesday, October

    29, 13

  146. 146 Obj4 = riakc_obj:update_value(Obj3, <<"2">>). Tuesday, October 29, 13

  147. 147 Obj5 = riakc_obj:update_value(Obj3, <<"22">>). Tuesday, October 29, 13

  148. 148 > riakc_pb_socket:put(Socket, Obj4). ok Tuesday, October 29, 13

  149. 149 > riakc_pb_socket:put(Socket, Obj5). {error,<<"failed">>} Tuesday, October 29, 13

  150. 150 Your client may vary Tuesday, October 29, 13

  151. 151 Your client may vary We’re working on it Tuesday,

    October 29, 13

  152. Tech Preview 152 Tuesday, October 29, 13

  153. 153 No AAE syncing No 2i No stats Tuesday, October

    29, 13

  154. 154 Will be in 2.0 final Tuesday, October 29, 13

  155. Coming Soon 155 Tuesday, October 29, 13

  156. 156 Datatypes Multi-DC Lightweight Tx? Perf benchmarks Tuesday, October 29,

    13

  157. 157 Datatypes Multi-DC Lightweight Tx? Perf benchmarks Tuesday, October 29,

    13

  158. 158 Datatypes Multi-DC Lightweight Tx? Perf benchmarks Tuesday, October 29,

    13

  159. 159 Datatypes Multi-DC Lightweight Tx? Perf benchmarks Tuesday, October 29,

    13

  160. 160 Datatypes Multi-DC Lightweight Tx? Perf benchmarks Tuesday, October 29,

    13

  161. Questions? 161 Tuesday, October 29, 13