For comprehensive error documentation, see
docs/ERRORS.md in the source repository.Startup & Installation
Virtualenv missing
Virtualenv missing
Error: No Or use the built-in setup command:
.venv directory found in backend/Solution:ModuleNotFoundError / No module named 'X'
ModuleNotFoundError / No module named 'X'
Error: Python dependency not installedSolution:
Port 8010 already in use
Port 8010 already in use
Error: Another process is using the backend portSolution:
Restart Asta to free and restart the port:Alternatively, set a different port in your environment:
Directory not found: backend
Directory not found: backend
Error: Running commands from the wrong directorySolution:
Always run
asta.sh commands from the asta/ repository root directory.Python Version Issues
Python 3.14 not supported
Python 3.14 not supported
Problem: Asta requires Python 3.12 or 3.13. Python 3.14 is not yet supported due to ChromaDB and pydantic compatibility issues.Solution:
Install a compatible Python version:The
asta.sh script automatically prefers Python 3.12 or 3.13 and will skip 3.14+ versions.API & Connection
API off / Cannot reach Asta API
API off / Cannot reach Asta API
Error: Backend is not runningSolution:Or run uvicorn manually:
CORS error
CORS error
Error: Request origin not allowedSolution:
Add your origin to Restart the backend after making changes:
ASTA_CORS_ORIGINS in backend/.env (comma-separated):Database & Persistence
SQLite lock errors
SQLite lock errors
Error: Database is locked or corruptSymptoms:Ensure the file is writable by your user.
database is lockederrorsResourceWarning: Connection deleted before being closed- Unable to save data
- Restart Asta to release locks:
- Check file permissions:
-
Release stale lock holders:
The
asta.sh restartcommand automatically attempts to release stale processes holding SQLite locks. - Start fresh (last resort): If the database is corrupt, back it up and remove it:
USER.md not found / empty
USER.md not found / empty
Error: Workspace context file not createdSolution:
Create or use the existing Or use the doc command with auto-fix:
workspace/USER.md file:AI Providers
No AI provider found
No AI provider found
Error: No API key configuredSolution:
Add at least one AI provider API key in Settings → API keys:
- Groq
- Gemini
- Claude (Anthropic)
- OpenAI
- OpenRouter
- Or run Ollama locally
Invalid or expired API key
Invalid or expired API key
Error: API key is wrong or expiredSolution:
- Update the key in Settings → API keys
- Or update
backend/.envdirectly - Restart the backend:
Client error 404 for Ollama
Client error 404 for Ollama
Error: Ollama not running or wrong URLSolution:
- Start Ollama:
- If Ollama is on another host, set the correct URL in
backend/.env:
- Restart the backend:
Learning / RAG
RAG not available / Ollama not available
RAG not available / Ollama not available
Error: RAG uses Ollama for embeddings; Ollama is not running or the model is missingSolution:
- Install Ollama:
- Pull the embedding model:
-
If Ollama is on another host, set
OLLAMA_BASE_URLinbackend/.env - Refresh the Learning page in Asta
RAG store failed
RAG store failed
Error: ChromaDB or FTS DB could not be createdSolution:
- Check disk space:
- Check write permissions:
- Set custom paths in
backend/.envif needed:
Logs & Debugging
Backend logs location:See Also
- Diagnostics - Run system diagnostics and health checks
- Source ERRORS.md - Complete error reference