GraphQL over REST
2 likes•560 views
The document discusses the evolution of API design from early standards like XML-RPC and SOAP to modern standards like REST and GraphQL. It outlines some limitations of REST APIs like underfetching and overfetching data through multiple requests. GraphQL is presented as a new API standard that was developed by Facebook to address these issues by allowing clients to fetch data from multiple resources through a single endpoint and request. Key features of GraphQL like declarative data fetching and a strong type system are also summarized.
GraphQL over REST
- 1. The Evolution of API Design
- 2. API History New API Standard GraphQL over REST GraphQL Fundamentals
- 3. Application
- 11. Simple Object Access Protocol SOAP SOAP-ENV : Envelope SOAP-ENV: Header SOAP-ENV: Body
- 12. Simple Object Access Protocol SOAP
- 14. REST
- 15. • Need for efficient data loading
- 16. • Need for efficient data loading • Variety of different frontend platforms
- 17. • Need for efficient data loading • Variety of different frontend platforms • Fast & Iterative development
- 22. With GraphQL, you model your business domain as a graph
- 24. A Query Language for your APIQuery Language
- 26. • New API Standard invented by Facebook
- 27. • New API Standard invented by Facebook • Presented publicly at React.js Conf 2015
- 28. • New API Standard invented by Facebook • Presented publicly at React.js Conf 2015 • Declarative Data Fetching
- 30. • New API Standard invented by Facebook • Presented publicly at React.js Conf 2015 • Declarative Data Fetching • Single Endpoint, Single Request
- 32. • New API Standard invented by Facebook • Presented publicly at React.js Conf 2015 • Declarative Data Fetching • Single Endpoint, Single Request • Strong Type System
- 38. Open Source License BSD + Patents
- 39. Open Source License BSD + Patents
- 40. Open Source License BSD + Patents
- 41. Open Source License OWFa v1.0
- 49. User Name
- 51. User Name Post Title Follower Name
- 59. /graphql
- 60. /graphql Fetch everything with a single request
- 61. /graphql Fetch everything with a single request
- 62. • Underfetching (N + 1 requests problem)
- 64. /graphql
- 65. • Underfetching (N + 1 requests problem) • Nested CompletionHandler
- 66. • Underfetching (N + 1 requests problem) • Nested CompletionHandler
- 67. • Underfetching (N + 1 requests problem) • Nested CompletionHandler • Overfetching
- 68. . . .
- 69. . . .
- 70. . . .
- 71. • Underfetching (N + 1 requests problem) • Nested CompletionHandler • Overfetching • Too Many Endpoints
- 74. • Underfetching (N + 1 requests problem) • Nested CompletionHandler • Overfetching • Too Many Endpoints • Versioning
- 78. • Underfetching (N + 1 requests problem) • Nested CompletionHandler • Overfetching • Too Many Endpoints • Versioning • Schemaless
- 80. • Underfetching (N + 1 requests problem) • Nested CompletionHandler • Overfetching • Too Many Endpoints • Versioning • Schemaless • Slow Product Development
- 93. User Requests (per hour) Authenticated 5000 Unauthenticated 60
- 94. User Requests (per hour) Authenticated 5000 Unauthenticated 60
- 95. User Requests (per hour) Authenticated 5000 Unauthenticated 60
- 96. • Node Limit Individual Call ≦ 500,000 total nodes
- 100. • Node Limit • Rate Limit 5,000 points per hour (points ≠ call)
- 102. 1
- 103. 1 100
- 104. 1 100 5000
- 105. 1 100 5000 5101
- 106. 1 100 5000 5101 / 100 = 51
- 113. Server
- 114. Server
- 115. Server Connection
- 116. Server Connection
- 117. Server Connection
- 120. Architectural Style HTTP, HTTPS Protocol Multiple Endpoints Schemaless Fixed Data Structure Versioning Standard API Transport Layer Agnostic Single Endpoint Strongly Typed Schema Flexible Data Structure Evolve API Without Versions