Troubleshooting
Something specific is broken and you just want it fixed - not another full tutorial to re-read. Find the symptom below.
Ollama & Models
Why Isn't Ollama Using My GPU?
nvidia-smi shows the card, but responses feel exactly as slow as CPU-only. Here's how to actually confirm what's happening, not just guess.
Ollama Says "Out of Memory" - Now What?
A model that should fit is failing to load or crashing mid-response. The fix depends on which kind of memory is actually the problem.
Ollama Service Won't Start
ollama run hangs or fails immediately, and the background service refuses to come up. Where to actually look before reinstalling anything.
Model Downloads Keep Failing or Coming Out Corrupted
A pull that "finishes" but the model behaves strangely, or repeatedly stalls at the same spot - what's actually happening and how to fix it.
Is 100% CPU Usage Normal for Ollama?
ollama ps shows the PROCESSOR column pegged at 100% CPU during a response. Usually yes, that's expected - here's how to tell if it actually isn't.
A Model Disappeared After an Update
A model you already pulled is just gone from ollama list. Almost always a path problem, not actual data loss - here's where to look.
Your Disk Suddenly Filled Up
Something is out of space right now and things are breaking. What to actually check and safely clean up, not just "delete some models."
Open WebUI & Docker
Ollama Is Running, But Open WebUI Can't Connect
The interface loads but shows no models, or errors trying to reach Ollama - even though ollama ps looks completely normal.
Open WebUI Logs In, But Chats Fail
The login screen works fine, but sending a message errors out or hangs. A different problem than "no models show up" - here's the real cause.
Docker Container Keeps Restarting
docker ps shows "Restarting" instead of settling down. Here's how to actually find out why instead of guessing and re-running the install command.