Agora Token Webservice

Written in Golang, using Gin framework to create a RESTful webservice for generating user tokens for use with the Agora.io platform.

Agora Advanced Guide: Token Management.

One-Click Deployments

Railway	Render	Heroku

How to Run

Set the APP_ID, APP_CERTIFICATE and CORS_ALLOW_ORIGIN env variables.

cp .env.example .env

go run cmd/main.go

Without using .env, you can also set the environment variables as such:

APP_ID=app_id APP_CERTIFICATE=app_cert CORS_ALLOW_ORIGIN=allowed_origins go run cmd/main.go

The pre-compiled binaries are also available in releases.

Docker

#1. To build the container with app id and certificate:

docker build -t agora-token-service --build-arg APP_ID=$APP_ID --build-arg APP_CERTIFICATE=$APP_CERTIFICATE --build-arg CORS_ALLOW_ORIGIN=$ALLOWED_ORIGINS .

#2. Run the container

docker run agora-token-service

Note: for testing locally

docker run -p 8080:8080 agora-token-service

Makefile

Build and run the docker container using make. The Makefile simplifies the build and run process by reducing them to a single command. The Makefile uses the .env file to add the list of --build-arg's to the docker build command and executes the docker run command after the build is completed. To avoid unnecessary rebuilds of the token server, the Makefile sets a build_marker target to watch the dockerfile, .go source code, and .env file. This enables a single command to build and run the container that only rebuilds as needed.

#1. Set the APP_ID, and APP_CERTIFICATE as env variables. Optionaly SERVER_PORT and CORS_ALLOW_ORIGIN can also be set as env variables, but they not required and there are defaults if they are not detected.

cp .env.example .env

#2. Run make

make

Execute the cleanup target to force a rebuild:

make cleanup

Endpoints

Ping

endpoint structure

/ping

response:

{"message":"pong"}

getToken

The getToken API endpoint allows you to generate tokens for different functionalities of the application. This section provides guidelines on how to use the getToken endpoint using HTTP POST requests.

Endpoint URL

POST /getToken

Request Body

The request body should contain a JSON payload with the required parameters for generating the tokens.

The following are the supported token types along with their required parameters:

RTC Token:

To generate an RTC token for video conferencing, include the following parameters in the request body:

{
    "tokenType": "rtc",
    "channel": "your-channel-name",
    "role": "publisher",  // "publisher" or "subscriber"
    "uid": "your-uid",
    "expire": 3600 // optional: expiration time in seconds (default: 3600)
}

RTM Token:

To generate an RTM token for Real-Time Messaging, include the following parameters in the request body:

{
    "tokenType": "rtm",
    "uid": "your-uid",
    "channel": "test", // optional: passing channel gives streamchannel. wildcard "*" is an option.
    "expire": 3600 // optional: expiration time in seconds (default: 3600)
}

Chat Token:

To generate a chat token, include the following parameters in the request body:

{
    "tokenType": "chat",
    "uid": "your-uid", // optional: for generating a user-specific chat token
    "expire": 3600 // optional: expiration time in seconds (default: 3600)
}

Response

Upon successful generation of the token, the API will respond with an HTTP status code of 200 OK, and the response body will contain the token in a JSON key "token".

If there is an error during token generation or if the request parameters are invalid, the API will respond with an appropriate HTTP status code and an error message in the response body.

Sample Usage

Here's an example of how to use the getToken API endpoint with a POST request using cURL:

Request:

curl -X POST -H "Content-Type: application/json" -d '{
    "tokenType": "rtc",
    "channel": "my-video-channel",
    "role": "publisher",
    "uid": "user123",
    "expire": 3600
}' "https://your-api-domain.com/getToken"

Reponse:

{
  "token": "007hbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJpc3MiOiJhZG1pbiIsInN1YiI6InVzZXIxMjMiLCJpYXQiOjE2MzEwNTU4NzIsImV4cCI6MTYzMTA1OTQ3Mn0.3eJ-RGwIl2ANFbdv4SeHtWzGiv6PpC3i0UqXlHfsqEw"
}

Deprecated Methods

The following methods are deprecated but still operational. While they continue to work for backward compatibility, it is advised to refrain from using them in new implementations due to potential future removal or replacement with more efficient alternatives.

RTC Token

The rtc token endpoint requires a tokenType (uid || userAccount), channelName, and the user's uid (type varies based on tokenType). expiry(optional) Pass an integer to represent the token lifetime in seconds.

endpoint structure

/rtc/:channelName/:role/:tokenType/:rtcuid/?expiry=3600

response:

{"rtcToken":"007rtc-token-djfkaljdla"}

RTM Token

The rtm token endpoint requires the user's rtmuid. expiry(optional) Pass an integer to represent the privelege lifetime in seconds. endpoint structure

/rtm/:rtmuid/?expiry=3600

response:

{"rtmToken":"007rtm-token-djfkaljdla"}

RTM + RTC Tokens

The rte token endpoint generates both the rtc and rtm tokens with a single request. This endpoint requires a tokenType (uid || userAccount), channelName, the user's rtcuid (type varies String/Int based on tokenType) and rtmuid which is a String. Omitting rtmuid will assume it's the same as rtcuid. expiry(optional) Pass an integer to represent the token lifetime in seconds.

endpoint structure

/rte/:channelName/:role/:tokenType/:rtcuid/:rtmuid/?expiry=3600

response:

{
  "rtcToken":"007rtc-token-djfkaljdla",
  "rtmToken":"007rtm-token-djfkaljdla" 
}

Chat Tokens

endpoint structure

app privileges:

chat/app/?expiry=3600

user privileges:

/chat/account/:chatid/?expiry=3600

expiry is an optional parameter for both.

response:

{
  "chatToken":"007chat-token-djfkaljdla"
}

Contributions

Contributions are welcome, please test any changes to the Go code with the following command:

APP_ID=<YOUR_APP_ID> APP_CERTIFICATE=<YOUR_APP_CERT> go test -cover github.com/AgoraIO-Community/agora-token-service/service