Volcengine plugin error: token too long

[Nothing1024]

Self Checks

• [x] This is only for bug report, if you would like to ask a question, please head to Discussions.
• [x] I have searched for existing issues search for existing issues, including closed ones.
• [x] I confirm that I am using English to submit this report (我已阅读并同意 Language Policy).
• [x] [FOR CHINESE USERS] 请务必使用英文提交 Issue，否则会被关闭。谢谢！:)
• [x] Please do not modify this template :) and fill in all the required fields.

Dify version

1.3.1

Cloud or Self Hosted

Self Hosted (Docker)

Steps to reproduce

Request Configuration

• model : doubao-1.5-vision-pro
• api_url: offical
• input params: one image(file id format) and some string params (for type selection)
• output params: just text
• method: workflow API request

The whole step flow is

1. upload the file to dify and get the file id

• method: NextJS fetch method
• request url : http://{dify-server}/v1/files/upload

2. request the dify workflow API

• method: NextJS fetch method
• request url: http://{dify-server}/v1/workflows/run
• params: image_url (Dify's id splice)

3. get the response

The whole step is automatic. Every epoch the step will request 3 times and wait for 2 seconds.

✔️ Expected Behavior

When I request the workflow API, it should give me the text result.

❌ Actual Behavior

But Sometimes the request will be error, the fetch response content is

TypeError: fetch failed
  1 |
  2 | export const DifyLLMRequest  = async (type: string, file_type: string, id: string)=>{
> 3 |   const response = await fetch("http://{dify-site}/v1/workflows/run", {
    |                   ^
  4 |     method: "POST",
  5 |     headers: {
  6 |       "Content-Type": "application/json", {
  [cause]: [Error [HeadersTimeoutError]: Headers Timeout Error] 
{
    code: 'UND_ERR_HEADERS_TIMEOUT'
  }
}

The dashboard show nothing, just the ”FAILED“ status task. When I try to find the logs, the error step just showed "running" status, no other outputs.

Then I try to get the docker logs (docker-plugin_daemon-1), I found that the plugin is always killed, so when it restart, the related request will be down.

Related logs

2025/05/07 07:05:37 stdio.go:166: [ERROR]plugin langgenius/volcengine_maas:0.0.13 has an error on stdout: bufio.Scanner: token too long
2025/05/07 07:05:37 run.go:132: [ERROR]plugin langgenius/volcengine_maas:0.0.13 exited with error: signal: killed

The token in token too long maybe not my prompt token, because I found various size image also have the same error.