Set rate limit for Oauth2 tokens

このページは、まだ日本語ではご利用いただけません。翻訳中です。

You can limit the tokens created in the OAuth2 plugin by using the Rate Limiting Advanced (RLA) plugin.

Note: You can’t configure a limit on the number of tokens that are created in the OAuth2 plugin.

Prerequisites

• The JSON responses shown in this guide are formatted using jq to make them easier to read. While useful, this tool is not necessary to complete the tasks of this tutorial. If you want to use jq to format your responses, add | jq to the end of each command.
• Rate Limiting Advanced plugin. The advanced version provides additional features like support for the sliding window algorithm and advanced Redis support for greater performance. Ensure that you have installed Kong Gateway Enterprise.

Configure rate limit for Oauth2 tokens

1. Create a workspace named teamA:

   curl --request POST \
    --url http://localhost:8001/workspaces \
    --data "name=teamA"
   

   The results should look like this:

   {
     "meta": {
       "color": null,
       "thumbnail": null
     },
     "name": "teamA",
     "comment": null,
     "config": {
       "meta": null,
    },
     "updated_at": 1699524963,
     "created_at": 1699524963,
     "id": "9e829c45-f478-4d7e-8c81-e5c9072548f3"
   }
   

2. Create a service named httpbin:

   curl --request POST \
    --url http://localhost:8001/teamA/services \
    --data "name=httpbin" \
    --data "url=https://httpbin.konghq.com/anything"
   

   The results should look like this:

   {
     "updated_at": 1699525785,
     "tags": null,
     "ca_certificates": null,
     "id": "741903b5-d952-47fb-97d6-de48156b6603",
     "port": 80,
     "tls_verify": null,
     "name": "mockbin",
     "tls_verify_depth": null,
     "client_certificate": null,
     "protocol": "http",
     "enabled": true,
     "created_at": 1699525785,
     "host": "httpbin.konghq.com",
     "connect_timeout": 60000,
     "path": "/anything",
     "read_timeout": 60000,
     "retries": 5,
     "write_timeout": 60000
   }
   

3. Create a route for the service:

   curl --request POST \
    --url http://localhost:8001/teamA/services/httpbin/routes \
    --data "name=httpbin" \
    --data "paths=/httpbin"
   

   The results should look like this:

   {
     "updated_at": 1699526004,
     "tags": null,
     "sources": null,
     "destinations": null,
     "path_handling": "v0",
     "id": "ce5a18dc-e0e2-474a-b622-5352e3689770",
     "preserve_host": false,
     "hosts": null,
     "name": "httpbin",
     "regex_priority": 0,
     "service": {
       "id": "741903b5-d952-47fb-97d6-de48156b6603"
     },
     "methods": null,
     "request_buffering": true,
     "response_buffering": true,
     "created_at": 1699526004,
     "paths": [
       "/httpbin"
     ],
     "strip_path": true,
     "https_redirect_status_code": 426,
     "headers": null,
     "protocols": [
       "http",
       "https"
     ],
     "snis": null
   }  
   

4. Enable the OAuth2 plugin:

   curl --request POST \
     --url http://localhost:8001/teamA/services/mockbin/plugins \
     --data "name=oauth2" \
     --data "config.scopes=email" \
     --data "config.enable_authorization_code=true" \
     --data "config.enable_client_credentials=true"
   

   The results should look like this:

   {
     "updated_at": 1699526484,
     "tags": null,
     "enabled": true,
     "route": null,
     "created_at": 1699526484,
     "consumer_group": null,
     "id": "82749d6e-44d7-47d3-a8f6-a182a0ac16bd",
     "instance_name": null,
     "name": "oauth2",
     "protocols": [
       "grpc",
       "grpcs",
       "http",
       "https",
       "ws",
       "wss"
     ],
     "service": {
       "id": "741903b5-d952-47fb-97d6-de48156b6603"
     },
     "consumer": null,
     "config": {
       "mandatory_scope": false,
       "accept_http_if_already_terminated": false,
       "auth_header_name": "authorization",
       "anonymous": null,
       "pkce": "lax",
       "reuse_refresh_token": false,
       "enable_password_grant": false,
       "enable_client_credentials": true,
       "provision_key": "1zusPwfUozzk9ADpcuXDYTivSWondujz",
       "scopes": [
         "email"
       ],
       "enable_implicit_grant": false,
       "enable_authorization_code": true,
       "hide_credentials": false,
       "persistent_refresh_token": false,
       "global_credentials": false,
       "refresh_token_ttl": 1209600,
       "token_expiration": 7200
     },
     "ordering": null
   }
   

5. Create a consumer named user1:

   curl --request POST \
     --url http://localhost:8001/teamA/consumers \
     --data "username=user1"
   

   The results should look like this:

   {
     "username": "user1",
     "updated_at": 1699526969,
     "tags": null,
     "created_at": 1699526969,
     "custom_id": null,
     "username_lower": "user1",
     "type": 0,
     "id": "691d8afc-b098-4374-937b-b1c36c8ef084"
   }  
   

6. Create a key with the value 123 for the consumer user1:

   curl -i -X POST localhost:8001/teamA/consumers/user1/key-auth \
     --data "key=123"
   

   The results should look like this:

   {
      "consumer":{
         "id":"691d8afc-b098-4374-937b-b1c36c8ef084"
      },
      "ttl":null,
      "created_at":1699527247,
      "key":"123",
      "id":"233005b4-3d03-4618-9de9-e05887dd2055",
      "tags":null
   }
   

7. Provision new OAuth 2.0 credentials with client_id set to 456 and client_secret set to 789:

   curl -i -X POST http://localhost:8001/teamA/consumers/user1/oauth2 \
     --data "name=demo" \
     --data "client_id=456" \
     --data "client_secret=789" \
     --data "redirect_uris=http://mockbin.org/"
   

   The results should look like this:

   {
      "tags":null,
      "created_at":1699527585,
      "redirect_uris":[
         "http://mockbin.org/"
      ],
      "consumer":{
         "id":"691d8afc-b098-4374-937b-b1c36c8ef084"
      },
      "hash_secret":false,
      "name":"demo",
      "client_type":"confidential",
      "id":"77dec565-8b1b-4e60-8ad1-59fa1a6f280e",
      "client_id":"456",
      "client_secret":"789"
   }   
   

8. Get a token for the client credentials:

   curl --request POST -k \
     --url 'https://localhost:8443/mockbin/oauth2/token' \
     --data "client_id=456" \
     --data "client_secret=789" \
     --data "grant_type=client_credentials"
   

   The results should look like this:

   {
     "token_type": "bearer",
     "expires_in": 7200,
     "access_token": "k55RovztGnAHwl8xC07gfbuC345TEDRY"
   }
   

   Make a note of this access_token to send proxy requests to the service.

9. Create a service named Oauth with oauth2/token endpoint:

     curl --request POST \
    --url http://localhost:8001/teamA/services \
    --data "name=Oauth" \
    --data "retries=0" \
    --data "host=localhost" \
    --data "path=/mockbin/oauth2/token" \
    --data "port=8443" \
    --data "protocol=https"
   

   The results should look like this:

   {
     "updated_at": 1699528090,
     "tags": null,
     "ca_certificates": null,
     "id": "6cba3d64-0676-4e68-9644-54f6bb1ec103",
     "port": 8443,
     "tls_verify": null,
     "name": "Oauth",
     "tls_verify_depth": null,
     "client_certificate": null,
     "protocol": "https",
     "enabled": true,
     "created_at": 1699528090,
     "host": "localhost",
     "connect_timeout": 60000,
     "path": "/mockbin/oauth2/token",
     "read_timeout": 60000,
     "retries": 0,
     "write_timeout": 60000
   }   
   

10. Create a route named oauthRoute for the Oauth service:

    curl --request POST \
      --url http://localhost:8001/teamA/services/Oauth/routes \
      --data "name=oauthRoute" \
      --data "paths=/oauth"
    

    The results should look like this:

    {
      "updated_at": 1699528555,
      "tags": null,
      "sources": null,
      "destinations": null,
      "path_handling": "v0",
      "id": "34fe2dda-9c8d-4525-93a0-e139212ff9dd",
      "preserve_host": false,
      "hosts": null,
      "name": "oauthRoute",
      "regex_priority": 0,
      "service": {
        "id": "6cba3d64-0676-4e68-9644-54f6bb1ec103"
      },
      "methods": null,
      "request_buffering": true,
      "response_buffering": true,
      "created_at": 1699528555,
      "paths": [
        "/oauth"
      ],
      "strip_path": true,
      "https_redirect_status_code": 426,
      "headers": null,
      "protocols": [
        "http",
        "https"
      ],
      "snis": null
    }   
    

11. Enable the Rate Limiting Advanced plugin with a limit of generating three tokens in span of 30 seconds.

    curl --request POST \
      --url http://localhost:8001/teamA/services/mockbin/plugins \
      --data "name=rate-limiting-advanced" \
      --data "config.limit=3" \
      --data "config.window_size=30" \
      --data "config.sync_rate=-1" \
      --data "config.strategy=local"
    

    The results should look like this:

    {
      "updated_at": 1699528815,
      "tags": null,
      "enabled": true,
      "route": null,
      "id": "8e475f2b-6acd-42b4-b96f-7cdd3084a3e8",
      "consumer_group": null,
      "ordering": null,
      "instance_name": null,
      "name": "rate-limiting-advanced",
      "protocols": [
        "grpc",
        "grpcs",
        "http",
        "https"
      ],
      "service": {
        "id": "741903b5-d952-47fb-97d6-de48156b6603"
      },
      "consumer": null,
      "config": {
        "identifier": "consumer",
        "error_code": 429,
        "hide_client_headers": false,
        "retry_after_jitter_max": 0,
        "path": null,
        "error_message": "API rate limit exceeded",
        "strategy": "local",
        "dictionary_name": "kong_rate_limiting_counters",
        "redis": {
          "username": null,
          "sentinel_role": null,
          "sentinel_master": null,
          "port": null,
          "database": 0,
          "timeout": 2000,
          "cluster_addresses": null,
          "sentinel_username": null,
          "ssl": false,
          "keepalive_pool_size": 30,
          "sentinel_addresses": null,
          "keepalive_backlog": null,
          "host": null,
          "connect_timeout": null,
          "send_timeout": null,
          "read_timeout": null,
          "sentinel_password": null,
          "server_name": null,
          "ssl_verify": false,
          "password": null
        },
        "enforce_consumer_groups": false,
        "consumer_groups": null,
        "limit": [
          3
        ],
        "header_name": null,
        "window_size": [
          30
        ],
        "sync_rate": -1,
        "disable_penalty": false,
        "namespace": "DO3rmnwLJFq4ju7tCZtd6HkKAf9rGEDQ",
        "window_type": "sliding"
      },
      "created_at": 1699528815
    }   
    

12. Send a proxy request to the Oauth/token endpoint three times within a span of 30 seconds. Ensure that you replace <accesstoken> with the access token that you made a note of:

    curl -i -X POST 'http://localhost:8000/mockbin?apikey=123' \
    --header "authorization: Bearer <accesstoken>" \
    

    When you send a request for the fourth time, you should get a 429 response code, showing you that the last request was rate limited:

    HTTP/1.1 429 Too Many Requests
    Date: Thu, 09 Nov 2023 12:41:01 GMT
    Content-Type: application/json; charset=utf-8
    Connection: keep-alive
    Retry-After: 38
    RateLimit-Reset: 38
    RateLimit-Remaining: 0
    X-RateLimit-Limit-30: 3
    X-RateLimit-Remaining-30: 0
    RateLimit-Limit: 3
    Content-Length: 37
    X-Kong-Response-Latency: 5
    Server: kong/3.4.1.1-enterprise-edition
       
    {"message":"API rate limit exceeded"}