scarlet river
)I’m requesting a credit refund/adjustment for time billed while the pod was not usable due to what appears to be a platform/networking issue.
Time affected (local):
2026-04-02 12:24:34 → 2026-04-02 16:10:20 (I was actively trying to fix it during this entire window)
What happened:
DNS/network failure inside the pod
Installing/reinstalling fails because the container cannot resolve GitHub:
Error: Resolving github.com (github.com)... failed: Temporary failure in name resolution
wget: unable to resolve host address 'github.com'
ComfyUI web UI never becomes available
ComfyUI starts (python main.py --listen ...) but does not bind to port 8188:
ss -ltnp | grep 8188 shows nothing listening
curl -I http://127.0.0.1:8188/ → curl: (7) Failed to connect ... Connection refused
Logs stop after GPU init (“Using pytorch attention���) and never reach the usual Running on http://0.0.0.0:8188.
ComfyUI-Manager makes it worse
When enabled (/workspace/ComfyUI/custom_nodes/comfyui-manager), startup runs “Security scan” / dependency install and then hangs after:
Prestartup times for custom nodes: comfyui-manager
Disabling/renaming that directory changes behavior, but the environment remains broken due to the DNS failure.
Environment:
Image: runpod/pytorch:2.4.0-py3.11-cuda12.4.1-devel-ubuntu22.04 (also tried recreating the container/pod)
GPU: RTX 4090 (was detected correctly when ComfyUI started)
Request: Please refund/credit back the charges incurred during 2026-04-02 12:24:34 → 16:10:20, since the pod was effectively unusable due to DNS/networking and service accessibility issues.
If you need pod ID, job ID, or logs, tell me where to find them in the dashboard and I’ll provide them.
Thank you.
prime wraith
stuck locust