Uh-oh! There was an issue with the response. CURL error: Timeout was reached

[sandivuk]

Open Terminal docker with Open WebUI docker

I'm not sure if I'm doing something wrong but sometimes I need to do rather complex tasks that causes the LLM to generate a very long script to execute.

I get the error in the title after what seems like five minutes elapse when the LLM does "executing run_command". I have tried setting the var OPEN_TERMINAL_EXECUTE_TIMEOUT=600 but it didn't seem to have an effect.

Shorter commands execute fine. Sometimes I need to run larger slower models for more precise output but they are running into this timeout. I can expand the "executing run_command" and see that the LLM is actively generating text so it isn't hanging or breaking. It just hits some pre-defined timeout and has to stop.

[tjbck]

The issue you are experiencing is due to Open WebUI's internal HTTP client timeout, not Open Terminal's execution timeout. Open WebUI enforces a strict 5-minute (300 seconds) timeout on all external API requests (including requests to your LLM, like Ollama, and requests to tool servers like Open Terminal), which overrides your OPEN_TERMINAL_EXECUTE_TIMEOUT setting.

When using "larger slower models" that take over 5 minutes to generate the long script (as you see when you expand "executing run_command"), Open WebUI forcefully closes the connection to the LLM because it has hit this 300-second limit, resulting in the timeout error.

To fix this, you need to increase the AIOHTTP_CLIENT_TIMEOUT environment variable on your Open WebUI docker container to match or exceed your desired timeout.

Add the following environment variable to your Open WebUI docker run command or docker-compose file: AIOHTTP_CLIENT_TIMEOUT=600 (or higher if needed)

After restarting the Open WebUI container, it will wait up to 600 seconds for the LLM to generate the script and for Open Terminal to execute it.