Open-webui erase every time I reboot my spark

Since my nvidia sync was giving me error port 11000 and unable to do custom port 12000 configuration for nvidia sync to access webui.

Instead I follow the instruction for the setup manually my open webui, but when I reboot my spark and reopen webui I got 500: internal error, I check docker ps on terminal the command return blank if like I never install docker and open-webui.

i test it out with two different command I found on the forum just to make sure I was not making a mistake. Both command erase every time i reboot my spark.

docker run -d -p 8080:8080 --gpus=all
-v open-webui:/app/backend/data
-v open-webui-ollama:/root/.ollama
–name open-webui ghcr.io/open-webui/open-webui:ollama

export NAME=open-webui
export IMAGE=“ghcr.io/open-webui/open-webui:ollama”
docker stop open-webui
docker rm open-webui
docker run -d -p 12000:8080 --gpus=all -v open-webui:/app/backend/data -v open-webui-ollama:/root/.ollama --name “${NAME}” “${IMAGE}” >/dev/null

well, the docker rm will delete your container. If you docker run --restart=unless-stopped … then the same container will automatically start whenever the machine reboots, instead of deleting the old one and starting a new one as your script is doing.

What is the exact output you get when you try running docker ps?

here is the result from docker ps :

docker ps
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES

Screenshot from 2026-01-21 20-39-56

thanks for the feedback but i did not put that part docker rm and my container still delete after reboot.

I understand, you want to have your container start on boot automatically right? I would recommend following coder543’s suggestion by using the --restart flag. You can find more details here: Start containers automatically | Docker Docs

what happen are I install open-webui by following the instruction but every time i reboot my spark, I have to reinstall my webui again with the model. I don’t want to keep reinstall open-webui and docker everytime I open my spark. I guess I have to do something else beside the instruction mention.

The commands you mention in your original post are working as designed, in fact you only need the first one

docker run -d -p 8080:8080 --gpus=all
-v open-webui:/app/backend/data
-v open-webui-ollama:/root/.ollama
–name open-webui ghcr.io/open-webui/open-webui:ollama

Docker containers will automatically shut down when you reboot your machine. In order to start the container again you would have to run the above command again. However, the --restart=always flag will tell the docker service to start the container any time it sees it is not running, due to an error, system reboot, etc. I would recommend going through the Docker documentation to learn how containers work and how they can be customized.

should I run this command

docker run -d -p 8080:8080 --gpus=all
-v open-webui:/app/backend/data
-v open-webui-ollama:/root/.ollama
–name open-webui ghcr.io/open-webui/open-webui:ollama
–restart=always \

do you think that is the command i should run

docker run -d -p 8080:8080 --gpus=all
-v open-webui:/app/backend/data
-v open-webui-ollama:/root/.ollama
–name open-webui ghcr.io/open-webui/open-webui:ollama
–restart=always \

@learnerbs22 instead of invoking docker run start the container you already have used previously, before the reboot. Look at your containers with docker container list –all to show active and inactive containers. Then on boot just have the desired container, i.e. open-webui you already used, with docker container start command. This way all your data will be there!

that is the command i run and now it work on spark every time i reboot.

docker run -d -p 8080:8080 --gpus=all \

-v open-webui:/app/backend/data \

-v open-webui-ollama:/root/.ollama \

--name open-webui --restart always ghcr.io/open-webui/open-webui:ollama

Glad to hear it. Please reach out if you have any other issues