Skip to main content

Troubleshooting

This guide helps you resolve common issues when using DevoxxGenie.

Chat UI Not Loading / GPU Rendering Error​

Problem​

On certain Windows systems with specific GPU driver configurations, the DevoxxGenie chat UI may fail to load with an error related to Skiko/Direct3D:

UnsatisfiedLinkError: Failed to load Direct3D native library

This occurs because the graphics rendering engine (Skiko) cannot initialize hardware-accelerated rendering on some GPU/driver combinations.

Automatic Fix​

Current DevoxxGenie versions force Skiko software rendering on Windows during startup, so the chat UI should load without touching GPU-accelerated rendering.

Manual Workarounds​

If you still experience issues, use one of these methods:

Option 1: Restart the IDE​

  1. Close IntelliJ IDEA fully
  2. Start IntelliJ IDEA again

This ensures the Windows startup override is applied before any chat UI is created.

Option 2: VM Options (Alternative for older versions)​

Add the following system property to your IDE's VM options:

  1. Go to Help > Edit Custom VM Options...
  2. Add this line: 
    -Dskiko.renderApi=SOFTWARE
  3. Save the file
  4. Restart IntelliJ IDEA

Option 3: Update GPU Drivers​

In some cases, updating your GPU drivers resolves the compatibility issue:

After updating drivers, restart IntelliJ IDEA and try DevoxxGenie again.

Still Not Working?​

If none of the above solutions work:

  1. Check the IntelliJ IDEA log files for detailed error messages:

    • Help > Show Log in Explorer/Finder
    • Look for errors related to Skiko, Direct3D, or UnsatisfiedLinkError

  2. Report the issue on GitHub Issues with:

    • Your operating system and version
    • GPU model and driver version
    • IntelliJ IDEA version
    • The relevant error log excerpt

Ollama Connection Issues​

Problem​

DevoxxGenie cannot connect to Ollama or models don't appear in the dropdown.

Solutions​

  1. Verify Ollama is running:

    curl http://localhost:11434/api/tags

    If this fails, start Ollama: ollama serve

  2. Check the base URL in settings:

    • Default should be: http://localhost:11434
    • If using Docker or remote Ollama, adjust accordingly

  3. Firewall/Antivirus: Ensure port 11434 is not blocked

  4. Model not showing: Click Refresh Models in the provider dropdown

See the full Ollama setup guide for more details.

API Key Errors​

Problem​

Getting "Invalid API key" or "Authentication failed" errors with cloud providers.

Solutions​

  1. Verify the API key is correct:

    • Open DevoxxGenie settings
    • Navigate to LLM Providers
    • Select the provider and check the API key field
    • Re-enter the key if unsure

  2. Check key permissions: Some providers require specific permissions or the key may be expired

  3. Verify provider status: Check if the LLM service is experiencing outages

Slow Responses with Local Models​

Problem​

Local LLM responses are very slow or timeout.

Solutions​

  1. Use a smaller model: Try llama3.2:3b or qwen2.5-coder:1.5b instead of larger models

  2. Check system resources:

    • Monitor CPU/GPU usage during generation
    • Ensure sufficient RAM is available
    • Close unnecessary applications

  3. Enable GPU acceleration for Ollama:

    ollama run llama3.2 --gpu

See Ollama performance tips for more optimization suggestions.

Plugin Not Visible After Installation​

Problem​

The DevoxxGenie icon doesn't appear in the IDE after installation.

Solutions​

  1. Restart IntelliJ IDEA — required after plugin installation

  2. Check if the plugin is enabled:

    • Settings > Plugins > Installed
    • Look for DevoxxGenie and ensure it's checked

  3. Invalidate caches:

    • File > Invalidate Caches / Restart...
    • Select Invalidate and Restart

  4. Verify IDE version: DevoxxGenie requires IntelliJ IDEA 2023.3.4 or later

Getting Help​

If your issue isn't covered here:

  1. Check the FAQ for common questions
  2. Search GitHub Issues for similar problems
  3. Join GitHub Discussions for community support
  4. Create a new issue with:
    • Clear description of the problem
    • Steps to reproduce
    • Expected vs actual behavior
    • Environment details (OS, IDE version, plugin version)
    • Relevant log excerpts