Skip to main content

FAQ

Question 1: Can I not use Ollama?

Answer: Yes

Configuration Steps

1. Change default model to Zhipu AI, disable Ollama service

# Disable Ollama service in docker-compose.yml
# Change default model to Zhipu AI
SPRING_AI_MODEL_CHAT: zhipuai
SPRING_AI_MODEL_EMBEDDING: zhipuai
# Disable Ollama service
SPRING_AI_OLLAMA_CHAT_ENABLED: false
SPRING_AI_OLLAMA_EMBEDDING_ENABLED: false

Or in source code configuration file:

# Change default model to Zhipu AI
spring.ai.model.chat=zhipuai
spring.ai.model.embedding=zhipuai
# Disable Ollama service
spring.ai.ollama.chat.enabled=false
spring.ai.ollama.embedding.enabled=false

2. Enable Zhipu AI service

Important Note

Must fill in real Zhipu AI API Key, otherwise cannot start normally. Using glm-4-flash model is free and will not incur costs.

Get API Key: Zhipu AI Console

# Configure in docker-compose.yml
SPRING_AI_ZHIPUAI_API_KEY: 'sk-xxx'
SPRING_AI_ZHIPUAI_CHAT_ENABLED: true
SPRING_AI_ZHIPUAI_CHAT_OPTIONS_MODEL: glm-4-flash
SPRING_AI_ZHIPUAI_CHAT_OPTIONS_TEMPERATURE: 0.7
SPRING_AI_ZHIPUAI_EMBEDDING_ENABLED: true

Or in source code configuration file:

spring.ai.zhipuai.api-key=sk-xxx
spring.ai.zhipuai.chat.enabled=true
spring.ai.zhipuai.chat.options.model=glm-4-flash
spring.ai.zhipuai.chat.options.temperature=0.7
spring.ai.zhipuai.embedding.enabled=true

faq_a

Common Questions

Q: If I don't want to use Zhipu AI, what if I want to use DeepSeek?

A: Since DeepSeek model doesn't support embedding, if you want to use knowledge base Q&A functionality, you must use at least one model that supports embedding. Therefore, when enabling DeepSeek, you must also enable Zhipu AI's embedding or Ollama models that support embedding.

Question 2: ChatClient dependency injection failure

Error message:

Unsatisfied dependency expressed through constructor parameter 0: No qualifying bean of type 'org.springframework.ai.chat.client ChatClient' available

faq1

Error Cause

Failed to connect to Ollama service, causing ChatClient injection to fail.

Solution

Modify Ollama service address configuration:

# Modify to actual Ollama service address in docker-compose.yml
SPRING_AI_OLLAMA_BASE_URL: http://host.docker.internal:11434

Or in source code configuration file:

spring.ai.ollama.base-url=http://127.0.0.1:11434

Question 3: Cannot find Protobuf classes

faq2

Error Analysis

Related Protobuf classes not generated.

Solution Steps

  1. Check Protobuf compilation tools

    protoc --version
    # Expected output: libprotoc 25.3
  2. Recompile project

    cd bytedesk  # Enter project root directory
    ./mvnw install -Dmaven.test.skip=true
Development Environment Recommendation

Recommended to use VS Code + Maven for development

Question 4: Message sorting error

faq3

Cause Analysis

Docker container timezone configuration error.

Solution

Fix Docker container timezone settings.

Question 5: Cannot get verification code or login failure

Problem Analysis

Cannot connect to server correctly, possibly due to incorrect server address configuration.

Solution Strategy

  1. Refresh page: Try refreshing browser page multiple times
  2. Switch server: If problem persists, try switching server

Operation Steps

Step 1: Click Settings

faq4

Step 2: Select Server

faq4

Step 3: Switch Server

faq4

Note

This method is only suitable for local testing, production environment is recommended to refer to Nginx Configuration

Question 6: How to use domain name access

Current Status

Default access address: http://serverIP:9003

Deployment Solutions

Method 2: Frontend-backend integrated deployment

Question 7: Upload functionality cannot work normally

Affected Scope

The following functions may not display normally:

  • Image upload
  • File upload
  • Knowledge base content

Solution

Modify upload address settings in configuration file.

Operation Method

In docker-compose.yaml or docker-compose-ollama.yaml file, replace 127.0.0.1 with actual server IP address or domain name:

# Please replace 127.0.0.1 with your server IP or domain name
BYTEDESK_UPLOAD_URL: http://127.0.0.1:9003
# Knowledge base access address, please modify to server actual address
BYTEDESK_KBASE_API_URL: http://127.0.0.1:9003
# Avatar access address, please modify to server actual address
BYTEDESK_FEATURES_AVATAR_BASE_URL: http://127.0.0.1:9003

Question 8: How to modify default password

Modification Method

Method 1: Modify through configuration file

Docker deployment:

Modify in docker-compose.yaml or docker-compose-ollama.yaml file:

BYTEDESK_ADMIN_EMAIL: admin@email.com
BYTEDESK_ADMIN_PASSWORD: admin

Source code deployment:

Modify in properties configuration file:

# Modify default admin password
bytedesk.admin.email=admin@email.com
bytedesk.admin.password=admin

Method 2: Modify after login

After logging into the system, modify password in personal profile page.

Question 9: Trial version prompt appears

Problem Phenomenon

Trial version prompt appears in the top right corner of admin backend:

faq_9

Cause

Using Enterprise Edition or Platform Edition features, but haven't completed paid authorization.

Solution

Method 1: Purchase official version

Refer to Purchase Guide to complete paid authorization.

Method 2: Switch to Community Edition

Modify configuration file to use Community Edition appkey

Question 10: How to customize brand information

Current Status

System uses Weiyu name and LOGO by default.

Customization Method

Refer to documentation: Customize Name and LOGO

Question 11: Default password for imported members

Default Settings

Default password for imported members: 123456

Modification Method

Source code configuration file

bytedesk.member.password=123456

Docker deployment configuration

Modify in docker-compose.yaml or docker-compose-ollama.yaml file:

BYTEDESK_MEMBER_PASSWORD: 123456

Question 12: How to close default demo page

Default Interface

faq_12_show_demo

Page After Closing

faq_12_hide_demo

Modification Method

Source code configuration file

bytedesk.custom.show-demo=false

Docker deployment configuration

Modify in docker-compose.yaml or docker-compose-ollama.yaml file:

BYTEDESK_CUSTOM_SHOW_DEMO: "false"

More Configuration

For detailed configuration instructions, please refer to: Member Configuration Documentation

Question 13: How to apply for licenseKey

How to Get

Log in to Weiyu Admin Backend with phone number, click Settings -> License Management to get trial version license.

faq_13

Fill the obtained licenseKey into configuration file

// properties configuration file
bytedesk.licenseKey=<KEY>

// docker-compose configuration file
BYTEDESK_LICENSE_KEY: <KEY>

For more details, refer to licenseKey Configuration

这页文档对您有帮助吗?