Tweet
Tweet

Slide 1

Slide 1 text

The Problem

Slide 2

Slide 2 text

No content

Slide 3

Slide 3 text

Server Node 1 Server Node 2

Slide 4

Slide 4 text

Node A Node B [User] [User] [User] [] [] [User] User connects User disconnects

Slide 5

Slide 5 text

There's no global time

Slide 6

Slide 6 text

Node A Node B [User] [User] [User] [] [] [User] User connects User disconnects

Slide 7

Slide 7 text

Node A Node B [User] [User] [User] [] [] [User] User connects User disconnects

Slide 8

Slide 8 text

Node A Node B

Slide 9

Slide 9 text

P.track(self, "users", "U1", %{})

Slide 10

Slide 10 text

P.track(self, "users", "U1", %{}) P.list("users") %{"U1"  %{metas: [%{phx_ref: "…"}]}}

Slide 11

Slide 11 text

P.track(self, "users", "U1", %{}) P.list("users") P.track(self, "users", "U2", %{}) %{"U1"  %{metas: [%{phx_ref: "…"}]}}

Slide 12

Slide 12 text

P.track(self, "users", "U1", %{}) P.list("users") P.track(self, "users", "U2", %{}) P.list("users") %{"U1"  %{metas: [%{phx_ref: "…"}]}, "U2"  %{metas: [%{phx_ref: "…"}]}} %{"U1"  %{metas: [%{phx_ref: "…"}]}}

Slide 13

Slide 13 text

P.track(self, "users", "U1", %{}) P.list("users") P.track(self, "users", "U2", %{}) P.list("users") %{"U1"  %{metas: [%{phx_ref: "…"}]}, "U2"  %{metas: [%{phx_ref: "…"}]}} P.track(self, "users", "U1", %{}) P.untrack(self, "users", "U1") %{"U1"  %{metas: [%{phx_ref: "…"}]}}

Slide 14

Slide 14 text

P.track(self, "users", "U1", %{}) P.list("users") P.track(self, "users", "U2", %{}) P.list("users") %{"U1"  %{metas: [%{phx_ref: "…"}]}, "U2"  %{metas: [%{phx_ref: "…"}]}} P.track(self, "users", "U1", %{}) P.untrack(self, "users", "U1") P.list("users") %{"U1"  %{metas: [%{phx_ref: "…"}]}, "U2"  %{metas: [%{phx_ref: "…"}]}} P.list("users") %{"U1"  %{metas: [%{phx_ref: "…"}]}, "U2"  %{metas: [%{phx_ref: "…"}]}} %{"U1"  %{metas: [%{phx_ref: "…"}]}}

Slide 15

Slide 15 text

CRDTs The science behind Phoenix Presence Maciej Kaszubowski

Slide 16

Slide 16 text

Conflict-free Replicated Data Type

Slide 17

Slide 17 text

Alternatives

Slide 18

Slide 18 text

• Single source of truth (DB) • Consensus algorithm • Resolving conflicts manually

Slide 19

Slide 19 text

Why CRDTs?

Slide 20

Slide 20 text

Eventually consistent Highly available Easy to use

Slide 21

Slide 21 text

Eventually consistent Highly available Easy to use Hard to create :(

Slide 22

Slide 22 text

Features

Slide 23

Slide 23 text

1. Commutative 2. Associative 3. Idempotent x y = y x (x y) z = x (y z) x x = x

Slide 24

Slide 24 text

Server Node 1 Server Node 2 Server Node 3

Slide 25

Slide 25 text

Client Client Client Client Client Client Client Client Client Server

Slide 26

Slide 26 text

Examples

Slide 27

Slide 27 text

Counters

Slide 28

Slide 28 text

Node A Node B +5 5 8 -2 3 +5 User connects User disconnects 0 0 5 3 8

Slide 29

Slide 29 text

Node A Node B +5 5 8 -2 8 +5 User connects User disconnects 0 0 5 3 13 10

Slide 30

Slide 30 text

G-Counter Grow-only counter

Slide 31

Slide 31 text

Node 2: 3 Value=8 Node 1: 5 Node 3: 1 Merge

Slide 32

Slide 32 text

Node 2: 3 Value=9 Node 1: 6 Node 3: 1 +1 Merge

Slide 33

Slide 33 text

PN-Counter Positive-Negative Counter

Slide 34

Slide 34 text

Node 2: P=2 N=2 Value=7 Node 1: P=5 N=2 Node 3: P=4 N=0 3 0 4 Merge

Slide 35

Slide 35 text

Node 2: P=2 N=2 Value=6 Node 1: P=5 N=3 Node 3: P=4 N=0 2 0 4 +1 Merge

Slide 36

Slide 36 text

Node 2: P=2 N=2 Value=8 Node 1: P=5 N=3 Node 3: P=6 N=0 2 0 6 +2 +1 Merge

Slide 37

Slide 37 text

Sets

Slide 38

Slide 38 text

Node A Node B [User] [User] [User] [] [] [User] User connects User disconnects

Slide 39

Slide 39 text

G-Set Grow-only set

Slide 40

Slide 40 text

Node 2: [1,2] Value=[1,2,3,4] Node 1: [1] Node 3: [3,4] Merge

Slide 41

Slide 41 text

Node 2: [1,2] Value=[1,2,3,4,5] Node 1: [1,5] Node 3: [3,4] Merge

Slide 42

Slide 42 text

2P-Set Two-phase set

Slide 43

Slide 43 text

Node 2: [1,2],[] Value=[1,2,3,4] Node 1: [1],[] Node 3: [3,4],[] Merge

Slide 44

Slide 44 text

Node 2: [1,2],[] Value=[1,2,3,4] Node 1: [1],[] Node 3: [3,4],[] Merge G-Set for adds

Slide 45

Slide 45 text

Node 2: [1,2],[] Value=[1,2,4] Node 1: [1],[] Node 3: [3,4],[3] Merge G-Set for removals

Slide 46

Slide 46 text

Elements cannot be re-added

Slide 47

Slide 47 text

Removes win

Slide 48

Slide 48 text

Node A Node B [ ] Add Remove [1] [ ] [1]

Slide 49

Slide 49 text

Node A Node B [ ] [1] Add Remove [1] [1] [ ] [1] [ ] [1]

Slide 50

Slide 50 text

Node A Node B [ ] [1] Add Remove [1] [1] [ ] [1] [ ] [1]

Slide 51

Slide 51 text

Node A Node B [ ] [1] [1] Add Remove [1] [1] [1] [ ] [1] [1] [1] [ ] [1]

Slide 52

Slide 52 text

OR-Set Observed-remove set

Slide 53

Slide 53 text

Node A Node B [ ] Add Remove [{A,1}] [ ] [{A,1}]

Slide 54

Slide 54 text

Node A Node B [ ] [ ] Add Remove [{A,1}] [{A,1}] [ ] [ ] [{A,1}] [{A,1}, ]

Slide 55

Slide 55 text

Node A Node B [ ] [ 1 ] Add Remove [{A,1}] [{A,1}] [ ] [ ] [{A,1}] [{A,1},{A,2}]

Slide 56

Slide 56 text

Node A Node B [ ] [ 1 ] Add Remove [{A,1}] [{A,1}] [ ] [ ] [{A,1}] [{A,1},{A,2}]

Slide 57

Slide 57 text

Node A Node B [ ] [ 1 ] Add Remove [{A,1}] [{A,1}] [ ] [ ] [{A,1}] [{A,1},{A,2}] [ 1 ] [{A,1},{A,2}] [{A,1},{A,2}] [ 1 ]

Slide 58

Slide 58 text

Node A Node B [ ] [ 1 ] Add Remove [{A,1}] [{A,1}] [ ] [ ] [{A,1}] [{A,1},{A,2}] [ 1 ] [{A,1},{A,2}] [{A,1},{A,2}] [ 1 ] [A] [A]

Slide 59

Slide 59 text

Add 1000 Remove 1000 Add Remove …

Slide 60

Slide 60 text

ORSWOT Observed-remove set without tombstones

Slide 61

Slide 61 text

No content

Slide 62

Slide 62 text

No content

Slide 63

Slide 63 text

Use Cases

Slide 64

Slide 64 text

No content

Slide 65

Slide 65 text

No content

Slide 66

Slide 66 text

No content

Slide 67

Slide 67 text

No content

Slide 68

Slide 68 text

• Load balancing / routing • Mobile clients synchronisation • Temporary data on the servers • Avoiding work duplication • Collaborative editing

Slide 69

Slide 69 text

No content

Slide 70

Slide 70 text

lasp-lang.readme.io

Slide 71

Slide 71 text

Summary

Slide 72

Slide 72 text

You're (almost) always designing a distributed system

Slide 73

Slide 73 text

Think about failures

Slide 74

Slide 74 text

Choose the correct tool for the job

Slide 75

Slide 75 text

References • https://medium.com/@istanbul_techie/a-look-at-conflict-free-replicated-data- types-crdt-221a5f629e7e • http://basho.com/posts/technical/distributed-data-types-riak-2-0/ • http://highscalability.com/blog/2014/10/13/how-league-of-legends-scaled- chat-to-70-million-players-it-t.html • https://hal.inria.fr/inria-00609399v1/document • https://developers.soundcloud.com/blog/roshi-a-crdt-system-for- timestamped-events

Slide 76

Slide 76 text

Thanks!