DataStax Enterprise Graph Fundamentals with Real World Example
1 like483 views
The document provides an overview of graph databases and their advantages over relational databases. It discusses how graph databases use vertices to represent entities and edges to represent relationships. Properties can be added to vertices and edges for additional metadata. Complex relationships can be easily modeled and queries performed using the Gremlin traversal language.
DataStax Enterprise Graph Fundamentals with Real World Example
- 1. DSE Graph Fundamentals Jon Haddad, DataStax Technical Evangelist @rustyrazorblade
- 2. © DataStax, All Rights Reserved. A Quick Look at Relational 2
- 3. © DataStax, All Rights Reserved. Tables • Very structured data • Relational calculus • Various normal forms 3 id name title country 1 jon evangelist usa 2 patrick chief evangelist usa 3 rebecca jr evangelist canada 4 billy ceo usa
- 4. © DataStax, All Rights Reserved. Selecting and Filtering 4 id name title country 1 jon evangelist usa 2 patrick chief evangelist usa 3 rebecca jr evangelist canada 4 billy ceo usa
- 5. © DataStax, All Rights Reserved. Selecting and Filtering 4 id name title country 1 jon evangelist usa 2 patrick chief evangelist usa 3 rebecca jr evangelist canada 4 billy ceo usa select * from people where country = 'canada';
- 6. © DataStax, All Rights Reserved. Selecting and Filtering 4 id name title country 1 jon evangelist usa 2 patrick chief evangelist usa 3 rebecca jr evangelist canada 4 billy ceo usa select * from people where country = 'canada';
- 7. © DataStax, All Rights Reserved. Foreign Keys 5 id name title country 1 jon evangelist 1 2 patrick chief evangelist 1 3 rebecca jr evangelist 2 4 billy ceo 1 id name 1 usa 2 canada 3 germany 4 japan
- 8. © DataStax, All Rights Reserved. Foreign Keys 5 id name title country 1 jon evangelist 1 2 patrick chief evangelist 1 3 rebecca jr evangelist 2 4 billy ceo 1 id name 1 usa 2 canada 3 germany 4 japan
- 9. © DataStax, All Rights Reserved. Foreign Keys 5 id name title country 1 jon evangelist 1 2 patrick chief evangelist 1 3 rebecca jr evangelist 2 4 billy ceo 1 id name 1 usa 2 canada 3 germany 4 japan
- 10. © DataStax, All Rights Reserved. Joins 6 select * from people join country on people.country = country.id; id name title country name 1 jon evangelist 1 usa 2 patrick chief evangelist 1 usa 3 rebecca jr evangelist 2 canada 4 billy ceo 1 usa
- 11. © DataStax, All Rights Reserved. Many to Many Relationships 7 id name title 1 jon evangelist 2 patrick chief evangelist 3 rebecca jr evangelist 4 billy ceo id name 1 usa 2 canada 3 germany 4 japan user country 1 1 1 2 2 1 3 2 4 1
- 12. © DataStax, All Rights Reserved. Many to Many Relationships 7 id name title 1 jon evangelist 2 patrick chief evangelist 3 rebecca jr evangelist 4 billy ceo id name 1 usa 2 canada 3 germany 4 japan user country 1 1 1 2 2 1 3 2 4 1
- 13. © DataStax, All Rights Reserved. Many to Many Relationships 7 id name title 1 jon evangelist 2 patrick chief evangelist 3 rebecca jr evangelist 4 billy ceo id name 1 usa 2 canada 3 germany 4 japan user country 1 1 1 2 2 1 3 2 4 1
- 14. © DataStax, All Rights Reserved. Many to Many Relationships 7 id name title 1 jon evangelist 2 patrick chief evangelist 3 rebecca jr evangelist 4 billy ceo id name 1 usa 2 canada 3 germany 4 japan user country 1 1 1 2 2 1 3 2 4 1
- 15. © DataStax, All Rights Reserved. Many to Many Relationships 7 id name title 1 jon evangelist 2 patrick chief evangelist 3 rebecca jr evangelist 4 billy ceo id name 1 usa 2 canada 3 germany 4 japan user country 1 1 1 2 2 1 3 2 4 1
- 16. © DataStax, All Rights Reserved. Joining 8 select * from people join people_country on people.id = people_country.user join country on people_country.country = country.id; id name title country name 1 jon evangelist 1 usa 1 jon evangelist 2 canada 2 patrick chief evangelist 1 usa 3 rebecca jr evangelist 2 canada 4 billy ceo 1 usa
- 17. © DataStax, All Rights Reserved. Joining 8 select * from people join people_country on people.id = people_country.user join country on people_country.country = country.id; id name title country name 1 jon evangelist 1 usa 1 jon evangelist 2 canada 2 patrick chief evangelist 1 usa 3 rebecca jr evangelist 2 canada 4 billy ceo 1 usa
- 18. © DataStax, All Rights Reserved. A Fun, Real World Example 9
- 19. © DataStax, All Rights Reserved. Media Database 1. People - actors, directories, assistants 2. Movies, TV, episodes 10
- 20. © DataStax, All Rights Reserved. Media Database 1. People - actors, directories, assistants 2. Movies, TV, episodes 10 person id name dob
- 21. © DataStax, All Rights Reserved. Media Database 1. People - actors, directories, assistants 2. Movies, TV, episodes 10 movie id name year person id name dob
- 22. © DataStax, All Rights Reserved. Media Database 1. People - actors, directories, assistants 2. Movies, TV, episodes 10 movie id name year tv_show id name person id name dob
- 23. © DataStax, All Rights Reserved. Media Database 1. People - actors, directories, assistants 2. Movies, TV, episodes 10 movie id name year tv_show id name person id name dob actor_movie person movie role
- 24. © DataStax, All Rights Reserved. Media Database 1. People - actors, directories, assistants 2. Movies, TV, episodes 10 movie id name year tv_show id name person id name dob actor_movie person movie role actor_tv person tv_show dob
- 25. © DataStax, All Rights Reserved. Media Database 1. People - actors, directories, assistants 2. Movies, TV, episodes 10 movie id name year tv_show id name person id name dob actor_movie person movie role actor_tv person tv_show dob
- 26. © DataStax, All Rights Reserved. Media Database 1. People - actors, directories, assistants 2. Movies, TV, episodes 10 movie id name year tv_show id name person id name dob actor_movie person movie role actor_tv person tv_show dob
- 27. © DataStax, All Rights Reserved. Media Database 1. People - actors, directories, assistants 2. Movies, TV, episodes 10 movie id name year tv_show id name person id name dob actor_movie person movie role actor_tv person tv_show dob episode tv_show season episode
- 28. © DataStax, All Rights Reserved. Media Database 1. People - actors, directories, assistants 2. Movies, TV, episodes 10 movie id name year tv_show id name person id name dob actor_movie person movie role actor_tv person tv_show dob episode tv_show season episode actor_episode episode actor
- 29. © DataStax, All Rights Reserved. Media Database 1. People - actors, directories, assistants 2. Movies, TV, episodes 10 movie id name year tv_show id name person id name dob actor_movie person movie role actor_tv person tv_show dob episode tv_show season episode actor_episode episode actor
- 30. © DataStax, All Rights Reserved. Media Database 1. People - actors, directories, assistants 2. Movies, TV, episodes 10 movie id name year tv_show id name person id name dob actor_movie person movie role actor_tv person tv_show dob episode tv_show season episode actor_episode episode actor
- 31. © DataStax, All Rights Reserved. Problems • Every entity gets a table • Lots of many to many tables • Very rigid structure • Going from one to many requires a migration and new data model 11
- 32. © DataStax, All Rights Reserved. Problems • Every entity gets a table • Lots of many to many tables • Very rigid structure • Going from one to many requires a migration and new data model 11
- 33. © DataStax, All Rights Reserved. How is this Solved with a Graph? 12
- 34. © DataStax, All Rights Reserved. Vertex • A vertex represents a "thing" • For example: Movie, Person 13
- 35. © DataStax, All Rights Reserved. Vertex • A vertex represents a "thing" • For example: Movie, Person 13 JCVD
- 36. © DataStax, All Rights Reserved. Vertex • A vertex represents a "thing" • For example: Movie, Person 13 JCVD Time cop
- 37. © DataStax, All Rights Reserved. JCVD Time cop Edge • Edges are labeled relationships • Edges have direction 14
- 38. © DataStax, All Rights Reserved. JCVD Time cop Edge • Edges are labeled relationships • Edges have direction 14 ActedIn
- 39. © DataStax, All Rights Reserved. Edges are always many to many 15 JCVD Time cop Blood sport
- 40. © DataStax, All Rights Reserved. Edges are always many to many 15 JCVD Time copActedIn Blood sport
- 41. © DataStax, All Rights Reserved. Edges are always many to many 15 JCVD Time copActedIn Blood sport ActedIn
- 42. © DataStax, All Rights Reserved. Edges are always many to many 15 JCVD Time copActedIn Blood sport ActedIn Directed
- 43. © DataStax, All Rights Reserved. The power of graph? 16
- 44. © DataStax, All Rights Reserved. Relationships 17
- 45. © DataStax, All Rights Reserved. Let's build on that 18
- 46. © DataStax, All Rights Reserved. Properties • Similar to fields in a table • Much more flexible • Meta properties - properties on properties • Can be on an edge or a vertex • Special property - a label 19 JCVD status amazing charm infinite odds of rendezvous 99.6% enemies decapitated 108747
- 47. © DataStax, All Rights Reserved. Summary 1. We don't define tables 2. We create vertices to represent real world objects 3. Relationships don't need to be explicitly modeled 4. Expressing complex relationships is not hard 5. Adding new edge types is easy 20
- 48. © DataStax, All Rights Reserved. Tinkerpop 3 & Gremlin • Cross database graph query framework • API for graph 21
- 49. © DataStax, All Rights Reserved. TinkerPop Basics 22
- 50. © DataStax, All Rights Reserved. Create a Vertex 23 graph.createVertex(label, "person", "name", "Jean Claude Van Damme") graph.createVertex(label, "movie", "name", "Bloodsport")
- 51. © DataStax, All Rights Reserved. Traversals 24 g = graph.traversal() g.V()
- 52. © DataStax, All Rights Reserved. Finding Vertices 25 g.V().has("person", "name", "Jean Claude van Damme")
- 53. © DataStax, All Rights Reserved. Finding Vertices 25 g.V().has("person", "name", "Jean Claude van Damme")
- 54. © DataStax, All Rights Reserved. Finding Vertices 25 g.V().has("person", "name", "Jean Claude van Damme")
- 55. © DataStax, All Rights Reserved. Finding Vertices 25 g.V().has("person", "name", "Jean Claude van Damme")
- 56. © DataStax, All Rights Reserved. Finding Vertices 25 g.V().has("person", "name", "Jean Claude van Damme")
- 57. © DataStax, All Rights Reserved. Create an Edge 26 g.V().has('person', 'name', "Jean Claude Van Damme").as('a'). V().has('movie', 'name', "Bloodsport").as('b'). addE("acted_in").property('year', 1988). from("a").to("b")
- 58. © DataStax, All Rights Reserved. Traversals 27 jcvd = g.V().has('person','name', "Jean Claude Van Damme")
- 59. © DataStax, All Rights Reserved. Traversals to neighbors 28 Bloodsport (1988) Kickboxer (1989) Timecop (1994)jcvd.out()
- 60. © DataStax, All Rights Reserved. Traversals to neighbors 28 Bloodsport (1988) Kickboxer (1989) Timecop (1994)jcvd.out()
- 61. © DataStax, All Rights Reserved. Traversals to neighbors 28 Bloodsport (1988) Kickboxer (1989) Timecop (1994)jcvd.out()
- 62. © DataStax, All Rights Reserved. Traversals to edges 29 Bloodsport (1988) Kickboxer (1989) Timecop (1994) jcvd.outE()
- 63. © DataStax, All Rights Reserved. Traversals to edges 29 Bloodsport (1988) Kickboxer (1989) Timecop (1994) jcvd.outE()
- 64. © DataStax, All Rights Reserved. Who cares? 30
- 65. © DataStax, All Rights Reserved. 31 SELECT movie.name from person JOIN roles on person.id = roles.person JOIN movies on roles.movie = movies.id WHERE person.name = "Jean Claude Van Damme"
- 66. © DataStax, All Rights Reserved. 32 jcvd.out("acted_in")
- 67. © DataStax, All Rights Reserved. 33
- 68. © DataStax, All Rights Reserved. 33 NO
- 69. © DataStax, All Rights Reserved. 34 jcvd.out("acted_in", "directed")
- 70. © DataStax, All Rights Reserved. 35 SELECT movies.name from person JOIN roles on person.id = roles.person JOIN movies on roles.movie = movies.id WHERE actor.name = "Jean Claude Van Damme" UNION SELECT movies.name from person JOIN directed on person.id = directed.person JOIN movies on directed.movie = movies.id WHERE actor.name = "Jean Claude Van Damme"
- 71. © DataStax, All Rights Reserved. SELECT movies.name from person JOIN roles on person.id = roles.person JOIN movies on roles.movie = movies.id WHERE actor.name = "Jean Claude Van Damme" UNION SELECT movies.name from person JOIN directed on person.id = directed.person JOIN movies on directed.movie = movies.id WHERE actor.name = "Jean Claude Van Damme" UNION SELECT movies.name from person JOIN produced on person.id = produced.person JOIN movies on produced.movie = movies.id WHERE actor.name = "Jean Claude Van Damme" UNION SELECT movies.name from person JOIN stage_hand on person.id = stage_hand.person JOIN movies on stage_hand.movie = movies.id
- 72. © DataStax, All Rights Reserved. SELECT movies.name from person JOIN directed on person.id = directed.person JOIN movies on directed.movie = movies.id WHERE actor.name = "Jean Claude Van Damme" UNION SELECT movies.name from person JOIN produced on person.id = produced.person JOIN movies on produced.movie = movies.id WHERE actor.name = "Jean Claude Van Damme" UNION SELECT movies.name from person JOIN stage_hand on person.id = stage_hand.person JOIN movies on stage_hand.movie = movies.id WHERE actor.name = "Jean Claude Van Damme" UNION SELECT movies.name from person JOIN stage_hand on person.id = stage_hand.person JOIN movies on stage_hand.movie = movies.id WHERE actor.name = "Jean Claude Van Damme"
- 73. © DataStax, All Rights Reserved. SELECT movies.name from person JOIN directed on person.id = directed.person JOIN movies on directed.movie = movies.id WHERE actor.name = "Jean Claude Van Damme" UNION SELECT movies.name from person JOIN produced on person.id = produced.person JOIN movies on produced.movie = movies.id WHERE actor.name = "Jean Claude Van Damme" UNION SELECT movies.name from person JOIN stage_hand on person.id = stage_hand.person JOIN movies on stage_hand.movie = movies.id WHERE actor.name = "Jean Claude Van Damme" UNION SELECT movies.name from person JOIN stage_hand on person.id = stage_hand.person JOIN movies on stage_hand.movie = movies.id WHERE actor.name = "Jean Claude Van Damme"
- 74. © DataStax, All Rights Reserved. 37 jcvd.out()
- 75. © DataStax, All Rights Reserved. Who acted with Van Damme and was born in Texas? 38
- 76. © DataStax, All Rights Reserved. Traversals 39 jcvd.out("acted_in") jcvd acted_in acted_in time cop bloodsport
- 77. © DataStax, All Rights Reserved. Traversals 39 jcvd.out("acted_in") jcvd acted_in acted_in .in("acted_in") time cop bloodsport
- 78. © DataStax, All Rights Reserved. Traversals 39 jcvd.out("acted_in") jcvd acted_in acted_in .in("acted_in") ron silveracted_in Bruce McGill acted_in Forest Whitaker acted_in New York Texas Texas time cop bloodsport
- 79. © DataStax, All Rights Reserved. Traversals 39 jcvd.out("acted_in") jcvd acted_in acted_in .in("acted_in") ron silveracted_in Bruce McGill acted_in Forest Whitaker acted_in New York Texas Texas time cop bloodsport
- 80. © DataStax, All Rights Reserved. Traversals 39 jcvd.out("acted_in") jcvd acted_in acted_in .has("state_of_birth", "Texas") .in("acted_in") ron silveracted_in Bruce McGill acted_in Forest Whitaker acted_in New York Texas Texas time cop bloodsport
- 81. © DataStax, All Rights Reserved. Traversals 39 jcvd.out("acted_in") jcvd acted_in acted_in .has("state_of_birth", "Texas") .in("acted_in") ron silveracted_in Bruce McGill acted_in Forest Whitaker acted_in New York Texas Texas time cop bloodsport
- 82. © DataStax, All Rights Reserved. Creating Indexes 40 schema.vertexLabel('person').index('by_year_born') .materialized().by('year_born').add() schema.vertexLabel('person').index('search'). search().by('biography').asText().add()
- 83. © DataStax, All Rights Reserved. Creating Indexes 40 schema.vertexLabel('person').index('by_year_born') .materialized().by('year_born').add() schema.vertexLabel('person').index('search'). search().by('biography').asText().add()
- 84. © DataStax, All Rights Reserved. Creating Indexes 40 schema.vertexLabel('person').index('by_year_born') .materialized().by('year_born').add() schema.vertexLabel('person').index('search'). search().by('biography').asText().add()
- 85. © DataStax, All Rights Reserved. Who has ANY relationship with any actor born in 1920? 41 g.V().has("person", "year_born", 1920). out("acted_in").in().hasLabel("person")
- 86. © DataStax, All Rights Reserved. index selection Who has ANY relationship with any actor born in 1920? 41 g.V().has("person", "year_born", 1920). out("acted_in").in().hasLabel("person")
- 87. © DataStax, All Rights Reserved. Who has ANY relationship with any actor born in 1920? 41 g.V().has("person", "year_born", 1920). out("acted_in").in().hasLabel("person") unconstrained relationships with no foreign keys
- 88. © DataStax, All Rights Reserved. How do I know Patrick? 42
- 89. © DataStax, All Rights Reserved. Path Finding! 43
- 90. © DataStax, All Rights Reserved. 44 jon = g.V().has("nickname", "rustyrazorblade") jon.until(has("nickname", "PatrickMcFadin")). repeat(both())
- 91. © DataStax, All Rights Reserved. My cousin went to school with Patrick 45 @rustyrazorblade @PatrickMcfadin
- 92. © DataStax, All Rights Reserved. My cousin went to school with Patrick 45 @rustyrazorblade @PatrickMcfadin cousin @oscar_the_grouch
- 93. © DataStax, All Rights Reserved. My cousin went to school with Patrick 45 @rustyrazorblade @PatrickMcfadin cousin @oscar_the_grouch attended Cal Poly
- 94. © DataStax, All Rights Reserved. My cousin went to school with Patrick 45 attended @rustyrazorblade @PatrickMcfadin cousin @oscar_the_grouch attended Cal Poly
- 95. © DataStax, All Rights Reserved. We have a mutual friend in the same company 46
- 96. © DataStax, All Rights Reserved. We have a mutual friend in the same company 46 friend
- 97. © DataStax, All Rights Reserved. We have a mutual friend in the same company 46 friend works_at
- 98. © DataStax, All Rights Reserved. We have a mutual friend in the same company 46 friend works_at works_at
- 99. © DataStax, All Rights Reserved. When do you use graph? 47
- 100. © DataStax, All Rights Reserved. Graph helps solves complex problems by utilizing of the power of relationships 48
- 101. © DataStax, All Rights Reserved. 49
- 102. © DataStax, All Rights Reserved. 49 Thank you!!!