Creating Apollo Server plugins
Extend Apollo Server with custom functionality
New in Apollo Server 3: All plugin lifecycle methods are async
, except for willResolveField
and schemaDidLoadOrUpdate
.
You can create your own Apollo Server plugins to perform custom operations in response to certain events. For example, a basic logging plugin might log the GraphQL query string associated with each request that's sent to Apollo Server.
The anatomy of a plugin
Plugins are JavaScript objects that implement one or more functions that respond to events. Here's a basic plugin that responds to the serverWillStart
event:
const myPlugin = {async serverWillStart() {console.log('Server starting up!');},};
If you're using TypeScript to create a plugin, the apollo-server-plugin-base
module exports the ApolloServerPlugin
interface for plugins to implement.
You can define a plugin in the same file where you initialize Apollo Server, or you can export it as a separate module:
module.exports = {async serverWillStart() {console.log('Server starting up!');},};
To create a plugin that accepts options, create a function that accepts an
options
object and returns a properly structured plugin object, like so:
module.exports = (options) => {return {async serverWillStart() {console.log(options.logMessage);},};};
Responding to events
A plugin specifies exactly which events it responds to by implementing functions that correspond to those events. The plugin in the examples above responds to the serverWillStart
event, which fires when Apollo Server is preparing to start up. Almost all plugin events are async
functions (i.e., functions that return Promise
s). The only exceptions
are willResolveField
and schemaDidLoadOrUpdate
.
A plugin can respond to any combination of supported events.
Responding to request lifecycle events
Plugins can respond to the following events associated with the GraphQL request lifecycle:
didResolveSource
parsingDidStart
validationDidStart
didResolveOperation
responseForOperation
executionDidStart
didEncounterErrors
willSendResponse
However, the way you define these functions is slightly different from the
serverWillStart
example above. First, your plugin must define the requestDidStart
function:
const myPlugin = {async requestDidStart() {console.log('Request started!');},};
The requestDidStart
event fires whenever Apollo Server receives a GraphQL request,
before any of the lifecycle events listed above. You can respond to this event
just like you respond to serverWillStart
, but you also use this function
to define responses for a request's lifecycle events, like so:
const myPlugin = {async requestDidStart(requestContext) {console.log('Request started!');return {async parsingDidStart(requestContext) {console.log('Parsing started!');},async validationDidStart(requestContext) {console.log('Validation started!');}}},};
As shown, the requestDidStart
function can optionally return an object that
defines functions that respond to request lifecycle events. This structure
organizes and encapsulates all of your plugin's request lifecycle logic, making it
easier to reason about.
Request lifecycle event flow
The following diagram illustrates the sequence of events that fire for each request. Each of these events is documented in Apollo Server plugin events.
Important: Any event below that can result in "Success" can also result in an error. Whenever an error occurs, the didEncounterErrors
event fires and the remainder of the "Success" path does not.
*The indicated events also support end hooks that are called when their associated step completes.
End hooks
Event handlers for the following events can optionally return a function that is invoked after the corresponding lifecycle phase ends:
(executionDidStart
returns an object containing an executionDidEnd
function instead of just a function as an end hook. This is because the returned object can also contain willResolveField
.)
Just like the event handlers themselves, these end hooks are async functions (except for the end hook for willResolveField
).
End hooks are passed any errors that occurred during the execution of that lifecycle phase. For example, the following plugin logs any errors that occur during any of the above lifecycle events:
const myPlugin = {async requestDidStart() {return {async parsingDidStart() {return async (err) => {if (err) {console.error(err);}}},async validationDidStart() {// This end hook is unique in that it can receive an array of errors,// which will contain every validation error that occurred.return async (errs) => {if (errs) {errs.forEach(err => console.error(err));}}},async executionDidStart() {return {async executionDidEnd(err) {if (err) {console.error(err);}}};},};},}
Note that the validationDidStart
end hook receives an array of errors that contains every validation error that occurred (if any). The willResolveField
end hook receives the error thrown by the resolver as the first argument and the result of the resolver as the second argument. The arguments to each end hook are documented in the type definitions in Request lifecycle events.
Inspecting request and response details
As the example above shows, requestDidStart
and request lifecycle functions accept a requestContext
parameter. This parameter is of type GraphQLRequestContext
, which includes a request
(of type GraphQLRequest
), along with a response
field (of type GraphQLResponse
) if it's available.
These types and their related subtypes are all defined in apollo-server-types/src/index.ts
.
Installing custom plugins
Add your plugins to Apollo Server by providing a plugins
configuration option to the ApolloServer
constructor, like so:
const { ApolloServer } = require('apollo-server');const ApolloServerOperationRegistry =require('@apollo/server-plugin-operation-registry');/* This example doesn't provide `typeDefs` or `resolvers`,both of which are required to start the server. */const { typeDefs, resolvers } = require('./separatelyDefined');const server = new ApolloServer({typeDefs,resolvers,csrfPrevention: true,cache: 'bounded',// You can import plugins or define them in-line, as shown:plugins: [/* This plugin is from a package that's imported above. */ApolloServerOperationRegistry({ /* options */ }),/* This plugin is imported in-place. */require('./localPluginModule'),/* This plugin is defined in-line. */{async serverWillStart() {console.log('Server starting up!');},}],})