Connection chaos or making sense of device discovery

I slightly increased the complexity of my Syncthing “cluster” by introducing an additional Syncthing instance on my homeserver. Now suddenly everyone wants to connect to everybody (well almost). This is a subset of the Syncthing instances which shows the problem:

I had two instances on two devices listening on 10.0.0.20:22000 (server) and 10.0.0.116:22000 (laptop). My router forwards external request to 22000 to 10.0.0.20:22000. I was used to the server complaining about connecting to itself, probably due to some global discovery problem (I think that’s even reported somewhere, but only visible in logs so not a problem).

Now there are two instances on server (:22000 and :22001) and still one on laptop (:22000). Router now forwards both 22000 and 22001 to 10.0.0.20. The instance on server :22000 is not (yet) connected to anybody. Still it gets New Device banners for the laptop instance and vice-versa.

What could be the cause for this? Getting rid of it is possible by ignoring the device, but this shouldn’t happen in the first place I think.

You should run syncthing on different ports on all devices behind lan and let global discovery expire (takes a few hours I think)

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.