Running custom docker images (used in Serverless) to use in Pods
Hey everyone - so here is my current situation:
- I have created a custom docker image for my serverless endpoint in Runpoint
- My local machine is a macbook so I am unable to execute the NVIDIA dependent comfyui installation I have in the image, so trying to see if I can run this on a Runpoint pod instead
- The use case is that I'm trying different workflows in ComfyUI that I want to test out in a Pod, before I deploy to the Serverless endpoint
I would love to hear what my options are. TIA!
6 Replies
You should be able to run the serverless worker on a pod by changing the START COMMAND in the template. Put in the path to the command to start Comfyui and it should run from pod.
Oh...I didn't think about running the serverless worker itself! Thank you!
So presumably, I just put the docker details in the container info and in start command, put the command
<path_to_comfyui>/python3 main.py ?
Sorry new to Pods and Docker, but how do I navigate the image to find the path to ComfyUI?
You can do one of the following (depending on shell that is installed on your image):
Once you have a shell inside the image you can issue the following command to find main.py
Hey @Encyrption - apologies for bring back up an old thread, but I'm working on this again and struggling 😮
How do I execute these commands on the shell in runpod?:
I created a pod template - and used the same docker image that i used for serverless (image 1)
I now have this pod running (image 2) - but unsure of what to do next?
you want to connect to a terminal?
use that connect button then web terminal
My example was not for using it after it is on RunPod it was showing you a way to get a shell from the image on your computer. For that you would pull the image to your computer and then run that command. You will need docker installed. But, that is just one way to get a shell. What @nerdylive has described should work after it is running on RP as a Pod.