GitHub issue middleware-manager #107, opened by
smoochyon GitHub
Summary
In my setup, Middleware Manager is configured to use standalone Traefik with active_data_source=traefik and no Pangolin present.
Codex found that GET /api/v1/traefik-config still appears to go through Pangolin fetch logic and returns 500 Internal Server Error when Pangolin is absent, even though Traefik itself is reachable and the loaded plugins are active.
This causes Traefik’s HTTP provider to stay in a retry loop and makes the Middleware Manager state misleading.
Environment
- Middleware Manager:
v4.5.0 - Traefik:
v3.6.12 - Deployment mode: standalone Traefik
- Active data source in config/UI:
traefik - Pangolin: not present in my setup
Traefik compose.yaml from my setup
services:
traefik:
container_name: traefik
image: traefik:v3.6.12@sha256:171c9c3565b29f6c133f1c1b43c5d4e5853415198e9e1078c001f8702ff66aec
logging:
driver: "json-file"
options:
max-file: "10"
max-size: "100m"
restart: unless-stopped
# Network
networks:
smoonet:
# Ports (host binding)
ports:
- "443:443/tcp" # HTTPS
- "444:444/tcp" # TCP (your custom)
- "8080:8080/tcp" # Dashboard
- "80:80/tcp" # HTTP
- "81:81/tcp" # Additional HTTP
# Volumes
volumes:
- /mnt/user/appdata/traefik:/etc/traefik # Config, certs, acme.json
# Environment variables
environment:
- CF_DNS_API_TOKEN=${CF_DNS_API_TOKEN} # …
---
💬 Discuss it here. Status, fixes and code review stay on [GitHub](https://github.com/hhftechnology/middleware-manager/issues/107).
<!-- hhf-forum-sync -->