Add comfyui service to docker-compose.yml, register with OmniRoute #41

Open
opened 2026-09-05 17:36:37 +00:00 by haylan · 1 comment
Owner

Part of #38

Question

Add a comfyui service to docker-compose.yml: ROCm-capable image, on the
ai-stack network (same pattern as llama-server), host port 8138 published
(for the future external nginx -> comfy.home route — nginx itself is outside
this stack, just the port mapping is needed here). Register it with
OmniRoute as a provider (same manual dashboard/API registration pattern used
for llama-server and searxng-search per docs/proxy-key-onboarding.md).
Verify a CLI request can reach ComfyUI's image-gen endpoint through
OmniRoute using an API key.

Part of #38 ## Question Add a comfyui service to docker-compose.yml: ROCm-capable image, on the ai-stack network (same pattern as llama-server), host port 8138 published (for the future external nginx -> comfy.home route — nginx itself is outside this stack, just the port mapping is needed here). Register it with OmniRoute as a provider (same manual dashboard/API registration pattern used for llama-server and searxng-search per docs/proxy-key-onboarding.md). Verify a CLI request can reach ComfyUI's image-gen endpoint through OmniRoute using an API key.
haylan added the wayfinder:task label 2026-09-05 17:36:37 +00:00
haylan self-assigned this 2026-09-05 19:56:31 +00:00
Author
Owner

Compose scaffolding done and pushed (ed83fca): comfyui service (yurisasc/comfyui-rocm7.1, gfx1201-tuned), ai-stack network, host port 8138 published, GID/UID auto-resolved in scripts/update.sh.

Remaining — needs to happen on the actual R9700 box, no shell access to it from this session:

  1. git pull && ./scripts/update.sh on the server.
  2. Register comfyui as an OmniRoute provider (base URL http://comfyui:8188) via the dashboard — same manual flow already used for llama-server, per docs/proxy-key-onboarding.md.
  3. Mint/reuse an API key and confirm a CLI request reaches ComfyUI's image-gen endpoint through OmniRoute.

Leaving open until that's confirmed — ping when done and I'll close it out.

Compose scaffolding done and pushed (ed83fca): comfyui service (yurisasc/comfyui-rocm7.1, gfx1201-tuned), ai-stack network, host port 8138 published, GID/UID auto-resolved in scripts/update.sh. Remaining — needs to happen on the actual R9700 box, no shell access to it from this session: 1. `git pull && ./scripts/update.sh` on the server. 2. Register comfyui as an OmniRoute provider (base URL `http://comfyui:8188`) via the dashboard — same manual flow already used for llama-server, per docs/proxy-key-onboarding.md. 3. Mint/reuse an API key and confirm a CLI request reaches ComfyUI's image-gen endpoint through OmniRoute. Leaving open until that's confirmed — ping when done and I'll close it out.
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: haylan/LLM-Server#41