GraphQL Stitching for Ruby

GraphQL stitching composes a single schema from multiple underlying GraphQL resources, then smartly proxies portions of incoming requests to their respective locations in dependency order and returns the merged results. This allows an entire location graph to be queried through one combined GraphQL surface area.

Supports:

• Merged object and abstract types.
• Multiple keys per merged type.
• Shared objects, fields, enums, and inputs across locations.
• Combining local and remote schemas.
• Type merging via arbitrary queries or federation _entities protocol.

NOT Supported:

• Computed fields (ie: federation-style @requires).
• Subscriptions, defer/stream.

This Ruby implementation is a sibling to GraphQL Tools (JS) and Bramble (Go), and its capabilities fall somewhere in between them. GraphQL stitching is similar in concept to Apollo Federation, though more generic. While Ruby is not the fastest language for a purely high-throughput API gateway, the opportunity here is for a Ruby application to stitch its local schemas together or onto remote sources without requiring an additional proxy service running in another language.

Getting started

Add to your Gemfile:

gem "graphql-stitching"

Run bundle install, then require unless running an autoloading framework (Rails, etc):

require "graphql/stitching"

Usage

The quickest way to start is to use the provided Client component that wraps a stitched graph in an executable workflow with caching hooks:

movies_schema = <<~GRAPHQL
  type Movie { id: ID! name: String! }
  type Query { movie(id: ID!): Movie }
GRAPHQL

showtimes_schema = <<~GRAPHQL
  type Showtime { id: ID! time: String! }
  type Query { showtime(id: ID!): Showtime }
GRAPHQL

client = GraphQL::Stitching::Client.new(locations: {
  movies: {
    schema: GraphQL::Schema.from_definition(movies_schema),
    executable: GraphQL::Stitching::HttpExecutable.new(url: "http://localhost:3000"),
  },
  showtimes: {
    schema: GraphQL::Schema.from_definition(showtimes_schema),
    executable: GraphQL::Stitching::HttpExecutable.new(url: "http://localhost:3001"),
  },
  my_local: {
    schema: MyLocal::GraphQL::Schema,
  },
})

result = client.execute(
  query: "query FetchFromAll($movieId:ID!, $showtimeId:ID!){
    movie(id:$movieId) { name }
    showtime(id:$showtimeId): { time }
    myLocalField
  }",
  variables: { "movieId" => "1", "showtimeId" => "2" },
  operation_name: "FetchFromAll"
)

Schemas provided in location settings may be class-based schemas with local resolvers (locally-executable schemas), or schemas built from SDL strings (schema definition language parsed using GraphQL::Schema.from_definition) and mapped to remote locations. See composer docs for more information on how schemas get merged.

While the Client constructor is an easy quick start, the library also has several discrete components that can be assembled into custom workflows:

• Composer - merges and validates many schemas into one supergraph.
• Supergraph - manages the combined schema, location routing maps, and executable resources. Can be exported, cached, and rehydrated.
• Request - prepares a requested GraphQL document and variables for stitching.
• Planner - builds a cacheable query plan for a request document.
• Executor - executes a query plan with given request variables.

Merged types

Object and Interface types may exist with different fields in different graph locations, and will get merged together in the combined schema.

To facilitate this merging of types, stitching must know how to cross-reference and fetch each variant of a type from its source location. This can be done using arbitrary queries or federation entities.

Merged types via arbitrary queries

Types can merge through arbitrary queries using the @stitch directive:

directive @stitch(key: String!) repeatable on FIELD_DEFINITION

This directive (or static configuration) is applied to root queries where a merged type may be accessed in each location, and a key argument specifies a field needed from other locations to be used as a query argument.

products_schema = <<~GRAPHQL
  directive @stitch(key: String!) repeatable on FIELD_DEFINITION

  type Product {
    id: ID!
    name: String!
  }

  type Query {
    product(id: ID!): Product @stitch(key: "id")
  }
GRAPHQL

shipping_schema = <<~GRAPHQL
  directive @stitch(key: String!) repeatable on FIELD_DEFINITION

  type Product {
    id: ID!
    weight: Float!
  }

  type Query {
    products(ids: [ID!]!): [Product]! @stitch(key: "id")
  }
GRAPHQL

client = GraphQL::Stitching::Client.new(locations: {
  products: {
    schema: GraphQL::Schema.from_definition(products_schema),
    executable:  GraphQL::Stitching::HttpExecutable.new(url: "http://localhost:3001"),
  },
  shipping: {
    schema: GraphQL::Schema.from_definition(shipping_schema),
    executable:  GraphQL::Stitching::HttpExecutable.new(url: "http://localhost:3002"),
  },
})

Focusing on the @stitch directive usage:

type Product {
  id: ID!
  name: String!
}
type Query {
  product(id: ID!): Product @stitch(key: "id")
}

• The @stitch directive is applied to a root query where the merged type may be accessed. The merged type identity is inferred from the field return.
• The key: "id" parameter indicates that an { id } must be selected from prior locations so it may be submitted as an argument to this query. The query argument used to send the key is inferred when possible (more on arguments later).

Each location that provides a unique variant of a type must provide one stitching query per key. The exception to this requirement are types that contain only a single key field:

type Product {
  id: ID!
}

The above representation of a Product type provides no unique data beyond a key that is available in other locations. Thus, this representation will never require an inbound request to fetch it, and its stitching query may be omitted. This pattern of providing key-only types is very common in stitching: it allows a foreign key to be represented as an object stub that may be enriched by data collected from other locations.

List queries

It's okay (even preferable in many circumstances) to provide a list accessor as a stitching query. The only requirement is that both the field argument and return type must be lists, and the query results are expected to be a mapped set with null holding the position of missing results.

type Query {
  products(ids: [ID!]!): [Product]! @stitch(key: "id")
}

# input:  ["1", "2", "3"]
# result: [{ id: "1" }, null, { id: "3" }]

See error handling tips for list queries.

Abstract queries

It's okay for stitching queries to be implemented through abstract types. An abstract query will provide access to all of its possible types. For interfaces, the key selection should match a field within the interface. For unions, all possible types must implement the key selection individually.

interface Node {
  id: ID!
}
type Product implements Node {
  id: ID!
  name: String!
}
type Query {
  nodes(ids: [ID!]!): [Node]! @stitch(key: "id")
}

Multiple query arguments

Stitching infers which argument to use for queries with a single argument. For queries that accept multiple arguments, the key must provide an argument mapping specified as "<arg>:<key>". Note the "id:id" key:

type Query {
  product(id: ID, upc: ID): Product @stitch(key: "id:id")
}

Multiple type keys

A type may exist in multiple locations across the graph using different keys, for example:

type Product { id:ID! }          # storefronts location
type Product { id:ID! upc:ID! }  # products location
type Product { upc:ID! }         # catelog location

In the above graph, the storefronts and catelog locations have different keys that join through an intermediary. This pattern is perfectly valid and resolvable as long as the intermediary provides stitching queries for each possible key:

type Product {
  id: ID!
  upc: ID!
}
type Query {
  productById(id: ID!): Product @stitch(key: "id")
  productByUpc(upc: ID!): Product @stitch(key: "upc")
}

The @stitch directive is also repeatable (requires graphql-ruby >= v2.0.15), allowing a single query to associate with multiple keys:

type Product {
  id: ID!
  upc: ID!
}
type Query {
  product(id: ID, upc: ID): Product @stitch(key: "id:id") @stitch(key: "upc:upc")
}

Class-based schemas

The @stitch directive can be added to class-based schemas with a directive class:

class StitchField < GraphQL::Schema::Directive
  graphql_name "stitch"
  locations FIELD_DEFINITION
  repeatable true
  argument :key, String, required: true
end

class Query < GraphQL::Schema::Object
  field :product, Product, null: false do
    directive StitchField, key: "id"
    argument :id, ID, required: true
  end
end

The @stitch directive can be exported from a class-based schema to an SDL string by calling schema.to_definition.

SDL-based schemas

A clean SDL string may also have stitching directives applied via static configuration by passing a stitch array in location settings:

sdl_string = <<~GRAPHQL
  type Product {
    id: ID!
    upc: ID!
  }
  type Query {
    productById(id: ID!): Product
    productByUpc(upc: ID!): Product
  }
GRAPHQL

supergraph = GraphQL::Stitching::Composer.new.perform({
  products:  {
    schema: GraphQL::Schema.from_definition(sdl_string),
    executable: ->() { ... },
    stitch: [
      { field_name: "productById", key: "id" },
      { field_name: "productByUpc", key: "upc" },
    ]
  },
  # ...
})

Custom directive names

The library is configured to use a @stitch directive by default. You may customize this by setting a new name during initialization:

GraphQL::Stitching.stitch_directive = "merge"

Merged types via Federation entities

The Apollo Federation specification defines a standard interface for accessing merged type variants across locations. Stitching can utilize a subset of this interface to facilitate basic type merging. The following spec is supported:

• @key(fields: "id") (repeatable) specifies a key field for an object type. The key fields argument may only contain one field selection.
• _Entity is a union type that must contain all types that implement a @key.
• _Any is a scalar that recieves raw JSON objects; each object representation contains a __typename and the type's key field.
• _entities(representations: [_Any!]!): [_Entity]! is a root query for local entity types.

The composer will automatcially detect and stitch schemas with an _entities query, for example:

accounts_schema = <<~GRAPHQL
  directive @key(fields: String!) repeatable on OBJECT

  type User @key(fields: "id") {
    id: ID!
    name: String!
    address: String!
  }

  union _Entity = User
  scalar _Any

  type Query {
    user(id: ID!): User
    _entities(representations: [_Any!]!): [_Entity]!
  }
GRAPHQL

comments_schema = <<~GRAPHQL
  directive @key(fields: String!) repeatable on OBJECT

  type User @key(fields: "id") {
    id: ID!
    comments: [String!]!
  }

  union _Entity = User
  scalar _Any

  type Query {
    _entities(representations: [_Any!]!): [_Entity]!
  }
GRAPHQL

client = GraphQL::Stitching::Client.new(locations: {
  accounts: {
    schema: GraphQL::Schema.from_definition(accounts_schema),
    executable: ...,
  },
  comments: {
    schema: GraphQL::Schema.from_definition(comments_schema),
    executable: ...,
  },
})

It's perfectly fine to mix and match schemas that implement an _entities query with schemas that implement @stitch directives; the protocols achieve the same result. Note that stitching is much simpler than Apollo Federation by design, and that Federation's advanced routing features (such as the @requires and @external directives) will not work with stitching.

Executables

An executable resource performs location-specific GraphQL requests. Executables may be GraphQL::Schema classes, or any object that responds to .call(location, source, variables, context) and returns a raw GraphQL response:

class MyExecutable
  def call(location, source, variables, context)
    # process a GraphQL request...
    return {
      "data" => { ... },
      "errors" => [ ... ],
    }
  end
end

A Supergraph is composed with executable resources provided for each location. Any location that omits the executable option will use the provided schema as its default executable:

supergraph = GraphQL::Stitching::Composer.new.perform({
  first: {
    schema: FirstSchema,
    # executable:^^^^^^ delegates to FirstSchema,
  },
  second: {
    schema: SecondSchema,
    executable: GraphQL::Stitching::HttpExecutable.new(url: "http://localhost:3001", headers: { ... }),
  },
  third: {
    schema: ThirdSchema,
    executable: MyExecutable.new,
  },
  fourth: {
    schema: FourthSchema,
    executable: ->(loc, query, vars, ctx) { ... },
  },
})

The GraphQL::Stitching::HttpExecutable class is provided as a simple executable wrapper around Net::HTTP.post. You should build your own executables to leverage your existing libraries and to add instrumentation. Note that you must manually assign all executables to a Supergraph when rehydrating it from cache (see docs).

Batching

The stitching executor automatically batches subgraph requests so that only one request is made per location per generation of data. This is done using batched queries that combine all data access for a given a location. For example:

query MyOperation_2 {
  _0_result: widgets(ids:["a","b","c"]) { ... } # << 3 Widget
  _1_0_result: sprocket(id:"x") { ... } # << 1 Sprocket
  _1_1_result: sprocket(id:"y") { ... } # << 1 Sprocket
  _1_2_result: sprocket(id:"z") { ... } # << 1 Sprocket
}

Tips:

• List queries (like the widgets selection above) are more compact for accessing multiple records, and are therefore preferable as stitching accessors.
• Assure that root field resolvers across your subgraph implement batching to anticipate cases like the three sprocket selections above.

Otherwise, there's no developer intervention necessary (or generally possible) to improve upon data access. Note that multiple generations of data may still force the executor to return to a previous location for more data.

Concurrency

The Executor component builds atop the Ruby fiber-based implementation of GraphQL::Dataloader. Non-blocking concurrency requires setting a fiber scheduler via Fiber.set_scheduler, see graphql-ruby docs. You may also need to build your own remote clients using corresponding HTTP libraries.

Additional topics

• Field selection routing
• Root selection routing
• Stitched errors
• Null results

Example

This repo includes a working example of several stitched schemas running across small Rack servers. Try running it:

bundle install
foreman start

Then visit the gateway service at http://localhost:3000 and try this query:

query {
  storefront(id: "1") {
    id
    products {
      upc
      name
      price
      manufacturer {
        name
        address
        products { upc name }
      }
    }
  }
}

The above query collects data from all locations, two of which are remote schemas and the third a local schema. The combined graph schema is also stitched in to provide introspection capabilities.

Tests

bundle install
bundle exec rake test [TEST=path/to/test.rb]