- 1. API with NestJS #1. Controllers, routing and the module structure
- 2. API with NestJS #2. Setting up a PostgreSQL database with TypeORM
- 3. API with NestJS #3. Authenticating users with bcrypt, Passport, JWT, and cookies
- 4. API with NestJS #4. Error handling and data validation
- 5. API with NestJS #5. Serializing the response with interceptors
- 6. API with NestJS #6. Looking into dependency injection and modules
- 7. API with NestJS #7. Creating relationships with Postgres and TypeORM
- 8. API with NestJS #8. Writing unit tests
- 9. API with NestJS #9. Testing services and controllers with integration tests
- 10. API with NestJS #10. Uploading public files to Amazon S3
- 11. API with NestJS #11. Managing private files with Amazon S3
- 12. API with NestJS #12. Introduction to Elasticsearch
- 13. API with NestJS #13. Implementing refresh tokens using JWT
- 14. API with NestJS #14. Improving performance of our Postgres database with indexes
- 15. API with NestJS #15. Defining transactions with PostgreSQL and TypeORM
- 16. API with NestJS #16. Using the array data type with PostgreSQL and TypeORM
- 17. API with NestJS #17. Offset and keyset pagination with PostgreSQL and TypeORM
- 18. API with NestJS #18. Exploring the idea of microservices
- 19. API with NestJS #19. Using RabbitMQ to communicate with microservices
- 20. API with NestJS #20. Communicating with microservices using the gRPC framework
- 21. API with NestJS #21. An introduction to CQRS
- 22. API with NestJS #22. Storing JSON with PostgreSQL and TypeORM
- 23. API with NestJS #23. Implementing in-memory cache to increase the performance
- 24. API with NestJS #24. Cache with Redis. Running the app in a Node.js cluster
- 25. API with NestJS #25. Sending scheduled emails with cron and Nodemailer
- 26. API with NestJS #26. Real-time chat with WebSockets
- 27. API with NestJS #27. Introduction to GraphQL. Queries, mutations, and authentication
- 28. API with NestJS #28. Dealing in the N + 1 problem in GraphQL
- 29. API with NestJS #29. Real-time updates with GraphQL subscriptions
- 30. API with NestJS #30. Scalar types in GraphQL
- 31. API with NestJS #31. Two-factor authentication
- 32. API with NestJS #32. Introduction to Prisma with PostgreSQL
- 33. API with NestJS #33. Managing PostgreSQL relationships with Prisma
- 34. API with NestJS #34. Handling CPU-intensive tasks with queues
- 35. API with NestJS #35. Using server-side sessions instead of JSON Web Tokens
- 36. API with NestJS #36. Introduction to Stripe with React
- 37. API with NestJS #37. Using Stripe to save credit cards for future use
- 38. API with NestJS #38. Setting up recurring payments via subscriptions with Stripe
- 39. API with NestJS #39. Reacting to Stripe events with webhooks
- 40. API with NestJS #40. Confirming the email address
- 41. API with NestJS #41. Verifying phone numbers and sending SMS messages with Twilio
- 42. API with NestJS #42. Authenticating users with Google
- 43. API with NestJS #43. Introduction to MongoDB
- 44. API with NestJS #44. Implementing relationships with MongoDB
- 45. API with NestJS #45. Virtual properties with MongoDB and Mongoose
- 46. API with NestJS #46. Managing transactions with MongoDB and Mongoose
- 47. API with NestJS #47. Implementing pagination with MongoDB and Mongoose
- 48. API with NestJS #48. Definining indexes with MongoDB and Mongoose
- 49. API with NestJS #49. Updating with PUT and PATCH with MongoDB and Mongoose
- 50. API with NestJS #50. Introduction to logging with the built-in logger and TypeORM
- 51. API with NestJS #51. Health checks with Terminus and Datadog
- 52. API with NestJS #52. Generating documentation with Compodoc and JSDoc
- 53. API with NestJS #53. Implementing soft deletes with PostgreSQL and TypeORM
- 54. API with NestJS #54. Storing files inside a PostgreSQL database
- 55. API with NestJS #55. Uploading files to the server
- 56. API with NestJS #56. Authorization with roles and claims
- 57. API with NestJS #57. Composing classes with the mixin pattern
- 58. API with NestJS #58. Using ETag to implement cache and save bandwidth
- 59. API with NestJS #59. Introduction to a monorepo with Lerna and Yarn workspaces
- 60. API with NestJS #60. The OpenAPI specification and Swagger
- 61. API with NestJS #61. Dealing with circular dependencies
- 62. API with NestJS #62. Introduction to MikroORM with PostgreSQL
- 63. API with NestJS #63. Relationships with PostgreSQL and MikroORM
- 64. API with NestJS #64. Transactions with PostgreSQL and MikroORM
- 65. API with NestJS #65. Implementing soft deletes using MikroORM and filters
- 66. API with NestJS #66. Improving PostgreSQL performance with indexes using MikroORM
- 67. API with NestJS #67. Migrating to TypeORM 0.3
- 68. API with NestJS #68. Interacting with the application through REPL
- 69. API with NestJS #69. Database migrations with TypeORM
- 70. API with NestJS #70. Defining dynamic modules
- 71. API with NestJS #71. Introduction to feature flags
- 72. API with NestJS #72. Working with PostgreSQL using raw SQL queries
- 73. API with NestJS #73. One-to-one relationships with raw SQL queries
- 74. API with NestJS #74. Designing many-to-one relationships using raw SQL queries
- 75. API with NestJS #75. Many-to-many relationships using raw SQL queries
- 76. API with NestJS #76. Working with transactions using raw SQL queries
- 77. API with NestJS #77. Offset and keyset pagination with raw SQL queries
- 78. API with NestJS #78. Generating statistics using aggregate functions in raw SQL
- 79. API with NestJS #79. Implementing searching with pattern matching and raw SQL
- 80. API with NestJS #80. Updating entities with PUT and PATCH using raw SQL queries
- 81. API with NestJS #81. Soft deletes with raw SQL queries
- 82. API with NestJS #82. Introduction to indexes with raw SQL queries
- 83. API with NestJS #83. Text search with tsvector and raw SQL
- 84. API with NestJS #84. Implementing filtering using subqueries with raw SQL
- 85. API with NestJS #85. Defining constraints with raw SQL
- 86. API with NestJS #86. Logging with the built-in logger when using raw SQL
- 87. API with NestJS #87. Writing unit tests in a project with raw SQL
- 88. API with NestJS #88. Testing a project with raw SQL using integration tests
- 89. API with NestJS #89. Replacing Express with Fastify
- 90. API with NestJS #90. Using various types of SQL joins
- 91. API with NestJS #91. Dockerizing a NestJS API with Docker Compose
- 92. API with NestJS #92. Increasing the developer experience with Docker Compose
- 93. API with NestJS #93. Deploying a NestJS app with Amazon ECS and RDS
- 94. API with NestJS #94. Deploying multiple instances on AWS with a load balancer
- 95. API with NestJS #95. CI/CD with Amazon ECS and GitHub Actions
- 96. API with NestJS #96. Running unit tests with CI/CD and GitHub Actions
- 97. API with NestJS #97. Introduction to managing logs with Amazon CloudWatch
- 98. API with NestJS #98. Health checks with Terminus and Amazon ECS
- 99. API with NestJS #99. Scaling the number of application instances with Amazon ECS
- 100. API with NestJS #100. The HTTPS protocol with Route 53 and AWS Certificate Manager
- 101. API with NestJS #101. Managing sensitive data using the AWS Secrets Manager
- 102. API with NestJS #102. Writing unit tests with Prisma
- 103. API with NestJS #103. Integration tests with Prisma
- 104. API with NestJS #104. Writing transactions with Prisma
- 105. API with NestJS #105. Implementing soft deletes with Prisma and middleware
- 106. API with NestJS #106. Improving performance through indexes with Prisma
- 107. API with NestJS #107. Offset and keyset pagination with Prisma
- 108. API with NestJS #108. Date and time with Prisma and PostgreSQL
- 109. API with NestJS #109. Arrays with PostgreSQL and Prisma
- 110. API with NestJS #110. Managing JSON data with PostgreSQL and Prisma
- 111. API with NestJS #111. Constraints with PostgreSQL and Prisma
- 112. API with NestJS #112. Serializing the response with Prisma
- 113. API with NestJS #113. Logging with Prisma
- 114. API with NestJS #114. Modifying data using PUT and PATCH methods with Prisma
- 115. API with NestJS #115. Database migrations with Prisma
- 116. API with NestJS #116. REST API versioning
- 117. API with NestJS #117. CORS – Cross-Origin Resource Sharing
- 118. API with NestJS #118. Uploading and streaming videos
- 119. API with NestJS #119. Type-safe SQL queries with Kysely and PostgreSQL
- 120. API with NestJS #120. One-to-one relationships with the Kysely query builder
- 121. API with NestJS #121. Many-to-one relationships with PostgreSQL and Kysely
The requirements of web applications constantly evolve, and so do the REST APIs they use. With the rise of the popularity of distributed systems and microservices, maintaining backward compatibility is increasingly important. Even if we maintain a monolithic architecture, there is a good chance that our frontend has a separate deployment pipeline. In this article, we implement REST API versioning in our NestJS application to handle updates gracefully and maintain a stable system.
Check out this repository to see the code from this article.
Versioning with NestJS
In one of the previous parts of this series, we changed the content property of the posts to paragraphs with a SQL migration.
1 2 3 4 5 6 7 8 |
ALTER TABLE "Post" ADD COLUMN "paragraphs" TEXT[]; UPDATE "Post" SET paragraphs = ARRAY[content]; ALTER TABLE "Post" DROP COLUMN content; |
If you want to know more about migrations with Prisma, check out API with NestJS #115. Database migrations with Prisma
If you use TypeORM, take a look at API with NestJS #69. Database migrations with TypeORM
Because of the above change, creating a new version of the endpoints related to posts makes sense. By implementing REST API versioning, we can maintain multiple versions of our controllers or individual routes. The most straightforward approach to API versioning is to include the version number in the URI of the request.
Adding versioning to an existing project
To start working with versions, we need to modify our bootstrap function.
main.ts
1 2 3 4 5 6 7 8 9 10 11 12 13 14 |
import { NestFactory } from '@nestjs/core'; import { AppModule } from './app.module'; import { VersioningType } from '@nestjs/common'; async function bootstrap() { const app = await NestFactory.create(AppModule); app.enableVersioning({ type: VersioningType.URI, }); await app.listen(3000); } bootstrap(); |
VersioningType.URI is the default value for the versioning type but we can provide it for the sake of readability.
Due to how we configured versioning with the app.enableVersioning method, our existing routes don’t include the version in the URL yet. Thanks to this approach, we are not introducing any breaking changes when starting to use versioning.
We now need to pass an additional argument to the @Controller() decorator that specifies the version.
posts.controller.ts
1 2 3 4 5 6 7 8 9 10 11 12 |
import { Controller } from '@nestjs/common'; import { PostsService } from './posts.service'; @Controller({ version: '2', path: 'posts', }) export default class PostsController { constructor(private readonly postsService: PostsService) {} // ... } |
Thanks to doing the above, we now have the /v2/ prefix in the URLs of endpoints related to posts.
To create a different version of the endpoints, we need to define a separate controller.
postsLegacy.controller.ts
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 |
import { ClassSerializerInterceptor, Controller, UseInterceptors, VERSION_NEUTRAL, } from '@nestjs/common'; import { PostsService } from './posts.service'; @Controller({ version: VERSION_NEUTRAL, path: 'posts', }) @UseInterceptors(ClassSerializerInterceptor) export default class PostsLegacyController { constructor(private readonly postsService: PostsService) {} // ... } |
By using VERSION_NEUTRAL as the version, NestJS uses the legacy controller when the version is not specified in the URL. Thanks to this approach, we can achieve backward compatibility if our project didn’t use versioning before.
We also need to remember about adding our new controller to the module.
posts.module.ts
1 2 3 4 5 6 7 8 9 10 11 12 |
import { Module } from '@nestjs/common'; import { PostsService } from './posts.service'; import PostsController from './posts.controller'; import { PrismaModule } from '../prisma/prisma.module'; import PostsLegacyController from './postsLegacy.controller'; @Module({ imports: [PrismaModule], controllers: [PostsController, PostsLegacyController], providers: [PostsService], }) export class PostsModule {} |
Adjusting our controller to the new schema
We now need to deal with the legacy API handling data using the old format with content instead of paragraphs. In one of the previous parts of this series, we learned how to serialize the response of our controller through a custom interceptor when using Prisma. To use it, we need to create a DTO class first.
postsResponseLegacy.dto.ts
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 |
import { Exclude, Expose } from 'class-transformer'; export class PostsResponseLegacyDto { id: number; title: string; authorId: number; scheduledDate: Date | null; @Exclude({ toPlainOnly: true, }) paragraphs: string[]; @Expose() get content() { return this.paragraphs.join('\n'); } } |
Above, we hide the paragraphs property that came from our database and create the content. Thanks to this, we can serve our data in a legacy format, even though the database stores it differently.
postsLegacy.controller.ts
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 |
import { ClassSerializerInterceptor, Controller, Get, Param, UseInterceptors, VERSION_NEUTRAL, } from '@nestjs/common'; import { PostsService } from './posts.service'; import { FindOneParams } from '../utils/findOneParams'; import { TransformDataInterceptor } from '../utils/transformData.interceptor'; import { PostsResponseLegacyDto } from './dto/legacy/postsResponseLegacy.dto'; @Controller({ version: VERSION_NEUTRAL, path: 'posts', }) @UseInterceptors(ClassSerializerInterceptor) export default class PostsLegacyController { constructor(private readonly postsService: PostsService) {} @Get(':id') @UseInterceptors(new TransformDataInterceptor(PostsResponseLegacyDto)) getPostById(@Param() { id }: FindOneParams) { return this.postsService.getPostById(id); } // ... } |
We can use a similar approach when creating entities by transforming the data posted by the user.
createPostLegacy.dto.ts
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 |
import { IsString, IsNotEmpty, IsNumber, IsOptional, IsISO8601, } from 'class-validator'; import { Exclude } from 'class-transformer'; export class CreatePostLegacyDto { @IsString() @IsNotEmpty() title: string; @IsOptional() @IsNumber({}, { each: true }) categoryIds?: number[]; @IsISO8601({ strict: true, }) @IsOptional() scheduledDate?: string; @Exclude({ toPlainOnly: true, }) @IsString() @IsNotEmpty() content: string; get paragraphs() { return [this.content]; } } |
We now need to use the above class in our legacy controller.
postsLegacy.controller.ts
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 |
import { Body, ClassSerializerInterceptor, Controller, Post, Req, UseGuards, UseInterceptors, VERSION_NEUTRAL, } from '@nestjs/common'; import { PostsService } from './posts.service'; import JwtAuthenticationGuard from '../authentication/jwt-authentication.guard'; import RequestWithUser from '../authentication/requestWithUser.interface'; import { TransformDataInterceptor } from '../utils/transformData.interceptor'; import { PostsResponseLegacyDto } from './dto/legacy/postsResponseLegacy.dto'; import { CreatePostLegacyDto } from './dto/legacy/createPostLegacy.dto'; @Controller({ version: VERSION_NEUTRAL, path: 'posts', }) @UseInterceptors(ClassSerializerInterceptor) export default class PostsLegacyController { constructor(private readonly postsService: PostsService) {} @Post() @UseGuards(JwtAuthenticationGuard) @UseInterceptors(new TransformDataInterceptor(PostsResponseLegacyDto)) async createPost( @Body() post: CreatePostLegacyDto, @Req() req: RequestWithUser, ) { return this.postsService.createPost(post, req.user); } // ... } |
Designing the API with versioning in mind
There is a high probability that we will need to add versioning to our application at some point. Since that’s the case, we might as well design it with versioning in mind.
main.ts
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 |
import { NestFactory } from '@nestjs/core'; import { AppModule } from './app.module'; import { VersioningType } from '@nestjs/common'; async function bootstrap() { const app = await NestFactory.create(AppModule); // ... app.enableVersioning({ type: VersioningType.URI, defaultVersion: '1', }); await app.listen(3000); } bootstrap(); |
By supplying defaultVersion: '1' to the enableVersioning method, we add the /v1/ prefix to all our endpoints. Now, whenever we create a new version for one of the controllers, the change in the URL feels more natural. For example, /v1/posts becomes /v2/posts.
The above also requires us to adjust our legacy controller. Instead of VERSION_NEUTRAL, we specify 1 as the version.
postLegacy.controller.ts
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 |
import { ClassSerializerInterceptor, Controller, UseInterceptors, } from '@nestjs/common'; import { PostsService } from './posts.service'; @Controller({ version: '1', path: 'posts', }) @UseInterceptors(ClassSerializerInterceptor) export default class PostsLegacyController { constructor(private readonly postsService: PostsService) {} // ... } |
Other types of versioning
While the URI versioning is default and very popular, NestJS also supports other types of versioning.
Header versioning
The header versioning uses a custom header to specify the version of the controller used to handle a particular request.
1 2 3 4 |
app.enableVersioning({ type: VersioningType.HEADER, header: 'Api-Version', }); |
Media type versioning
The Accept request header indicates which content types the API client can understand. A good example is Accept: application/json.
With media-type versioning, we use the Accept header to specify the version, such as Accept: application/json;v=2.
1 2 3 4 |
app.enableVersioning({ type: VersioningType.MEDIA_TYPE, key: 'v=', }); |
Custom versioning type
If none of the above solutions fit our API, we can implement a custom versioning.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 |
import { AppModule } from './app.module'; import { VersioningType } from '@nestjs/common'; import { isRecord } from './utils/isRecord'; async function bootstrap() { const app = await NestFactory.create(AppModule); app.enableVersioning({ type: VersioningType.CUSTOM, extractor: (request: unknown) => { if (isRecord(request) && isRecord(request.query)) { return String(request.query.version); } return '1'; }, }); // ... await app.listen(3000); } bootstrap(); |
Unfortunately, the type of the request is unknown and we need to use type assertions.
Summary
In this article, we’ve gone through the idea of REST API versioning. We used URI versioning to create a legacy version of our controller and transformed data to match the previous format our API was using. We also learned about other versioning types, such as header and media-type versioning.
API versioning can be a handy tool to preserve backward compatibility. However, we need to remember to remove the older versions of our API when they are not required anymore to avoid cluttering our codebase.