TCP Router - how to disable backend TCP services check


I configured Traefik TCP Router to serve an FTP service hosted on a Kubernetes cluster (with KubernetesCDR provider integration).

FTP server works in passive mode so I configured Traefik to manage and forward both control (21) and data port (passive ports range).
It works well, but I have a collateral issue.

FTP Data ports are normally closed and are opened dynamically only during the data transfer session.
Traefik service (I suppose a service health check process or similar) keeps trying to connect to data ports always. The results is a lot of connection errors and a lot of general error events (about 10 for second).

This is the Traefik event log sample:
{"level":"error","msg":"Error while connecting to backend: dial tcp connect: connection refused","time":"2023-02-03T07:25:25Z"}

At the FTP side I get similar error log for incomplete connection for ftp control port.

If it's a service check, is it possible to disable it or change check interval?
I've searched a lot but haven't found a way to do it.


I realized that the health-check wasn't triggered by Traefik, but by the load balanced in front of Traefik.

This topic was automatically closed 3 days after the last reply. New replies are no longer allowed.