GraphQL Basic Usage

(This section assumes you have read the section What is Coraza? )

• By default, GraphQL validation is not active on any route.
• GraphQL related settings are at spec.coraza.graphql and spec.routes[].coraza.graphql, plus general Coraza settings at spec.coraza and spec.routes[].coraza in the API Reference .
• To activate GraphQL on a route, you have to first define a config and then reference it at a route, see the above settings.
• The GraphQL schema can be provided from a ConfigMap or from an OCI image.
• You can turn GraphQL validation off and on in various ways, by disabling Coraza and/or GraphQL globally or per route, see the above settings.
• You can configure limits for complexity, depth and batch size, and allow introspection or not, also with the above settings.
• Regarding size limits, see also the section Large Payloads .
• To use auto-learning, it is recommended to first use GraphQL in mode DETECT, and then auto-learn using the Auto-Learning CLI with processing argument graphql.

Basic Example

Here's a basic example with GraphQL with default maxima for complexity, depth and batch size (and introspection not allowed by default), active in mode BLOCK on the route /anything, and some limits on request and response body, and with CRS disabled.

spec:
  [...]
  coraza:
    requestBodyAccess: true
    requestBodyLimitKb: 1024
    responseBodyAccess: false
    responseBodyLimitKb: 128
    crs:
      defaultEnabled: false
    graphql:
      configs:
      - name: "graphql-book"
        schemaSource:
          configMap: "graphql-book-config-map"
          key: "schema.graphql"
  routes:
    - match:
        path: /anything
        pathType: "PREFIX"
      coraza:
        graphql:
          enabled: true
          ref: "graphql-book"
          mode: "BLOCK"
      backend:
        address: httpbin
        port: 8000
        protocol:
          selection: h1