Docker deployment Dify + RAGFlow avoidance guide

After installing dify and RAGFlow in Docker, you may encounter the following problems:

1. Port conflict problem

1. Symptom : Dify and RAGFlow use the same ports (80 and 443) by default, causing a service to fail to start. Solution :
   

• Modify the port mapping of RAGFlow, for example docker-compose.yml Medium Adjustment:

  ports:
    - "8000:80" # Container port 80 is mapped to host port 8000   
    - "4333:443" # Container port 443 is mapped to host port 4333  
  

• Restart the service:docker-compose up -d.
• Check firewall rules to make sure new ports (such as 8000, 4333) are open.

2. Dependency Service Conflict

1. Phenomenon : Both rely on Redis, and conflicts will occur if the default configurations are the same. Solution :
   

• Modify the Redis configuration of RAGFlow (in .env in the document):

  redis_port=7379 # Change the default port
  redis_password=custom password
  

• Make sure Dify's Redis configuration is independent (eg use a different port or password).

3. Insufficient hardware resources

1. Symptom : RAGFlow has high resource requirements (CPU ≥ 4 cores, memory ≥ 16GB, disk ≥ 50GB). When resources are insufficient, it fails to start or runs slowly. Solution :
   

• Check your server configuration, upgrade your hardware, or adjust Docker resource allocation (set CPU/memory limits via Docker Desktop).
• For Linux systems, kernel parameters need to be set:

  sudo sysctl -w vm.max_map_count=262144   # Temporary effect
  sudo vim /etc/sysctl.conf                # Permanent effect
  

4. Model Integration Configuration Issues

1. Symptom : When integrating a local model (such as Ollama), the API is not authorized or the model cannot be loaded. Solution :
   

• Create an API Key in RAGFlow and record the repository ID.
• Dify .env Fill in the RAGFlow API address and Key in the file:

  API_ENDPOINT=http://<RAGFlow_IP>:9380/api/v1/dify
  API_KEY=your_ragflow_api_key
  

• Disable Dify's Rerank model and give priority to RAGFlow's parsing results.

5. Docker environment issues

1. Symptom (Windows specific):

• WSL2 is not enabled : Docker Desktop cannot be started. Solution : Run PowerShell with administrator privileges:
  

  dism.exe /online /enable-feature /featurename:Microsoft-Windows-Subsystem-Linux /all
  wsl --set-default-version 2
  

• Image pull failed : Network problems caused image download to time out. Solution : Use a domestic image source or a pre-packaged one-click installation package.
  

6. Document parsing and retrieval effect issues

1. Phenomenon : Dify's native knowledge base has poor parsing effect on PDF scans and tables. Solution :
   

• Upload files through RAGFlow and take advantage of its deep document parsing capabilities (supports OCR, layout recognition, etc.).
• In Dify, associate RAGFlow's external knowledge base and adjust parameters such as  TopK (number of retrievals) and  similarity threshold .

Post-launch verification steps

1. Access services :

• Dify:http://<IP>:80
• RAGFlow:http://<IP>:8000(If the port has been modified).

• Log troubleshooting :

  docker logs -f ragflow-server   # View RAGFlow logs
  docker logs dify-app           # View Dify logs
  

• Test the knowledge base : Create an application in Dify and associate it with the RAGFlow knowledge base to verify the accuracy of questions and answers.

---

Summary and recommendations

• Recommended deployment scheme :
• Deploy Dify and RAGFlow on the same host and communicate through the intranet IP to reduce latency.
• Using RAGFlow docker-compose-gpu.yml Configure to accelerate processing if GPU is available.
• Tips for avoiding pitfalls : Windows users need to first resolve the WSL2 and port conflicts; Linux users need to check kernel parameters and resource allocation.