Aulafy
Courses/Claude Code + Local AI/Open WebUI troubleshooting and lightweight production

Open WebUI troubleshooting and lightweight production

Open WebUI is very convenient for using local models, but when you turn it into a daily tool you need to think about versions, volumes, backups, UI changes, Ollama connectivity, and recovery.

  • Avoid losing chats, users, or configuration from a bad update.
  • Diagnose Ollama connectivity and tool issues.
  • Set up a lightweight installation for small business or small team use.
TerminalCode
# Ver contenedores y logs
docker ps
docker logs -f open-webui

# Comprobar Ollama desde el host
curl http://localhost:11434/api/tags

# Comprobar desde un contenedor en la misma red
docker exec -it open-webui sh
curl http://host.docker.internal:11434/api/tags

Pre-update checklist

  • Note the current image version.
  • Back up the Open WebUI volume.
  • Read release notes if you rely on tools, auth, or RAG.
  • Test on a copy if more people use it.
  • Have a rollback plan.
TerminalCode
# Backup simple de volumen Docker
docker run --rm \
  -v open-webui:/data \
  -v "$PWD/backups":/backup \
  alpine tar czf /backup/open-webui-$(date +%Y%m%d).tgz -C /data .

Minimal lightweight production

  • Open WebUI with persistent volume.
  • Ollama running separately and monitored.
  • Qdrant with its own volume if you use RAG.
  • Reverse proxy with HTTPS if it goes beyond localhost.
  • Tested backups, not just configured.

Official sources

  • Open WebUI documentation
  • Ollama API
  • Qdrant documentation