Messages go out, but remote messages don't show up (federation issues)

On my new instance, if I send a message out, it will get seen on other instances, at least if I mention someone on those instances.

Replying to those messages from those other instance, there are no messages that show up, or almost none. I have 2 non-local toots on my federated timeline, from two different instances. I’m hoping that indicates that it’s just that it will take a bit of time, but some feedback would be appreciated.

Any way that I can find a way that I can do some debugging on this? I’m not really sure where more detailed logs might be.

Unfortunately, there are a lot of mis-configurations that can cause federation to fail. It’s hard to debug specific issues without logs or a link to the instance in question. In practice, misconfigured ipv6 addresses are the root cause of 60% of servers that have trouble receiving posts from other instances.

As a first step, I would try following a couple users on other servers to try and see if the problem is consistent for a single domain.

Thank you, so far, I have taken a few different things to see what might be doing. I took my follow list on and imported it onto the one account that exists there, and it seems that has helped with some stuff.

It’s not consistent with regards to how an account gets handled on a given instance, but things are a bit better now.

The instance I have setup is and I’m not sure what I have for logs, or even what to look with to try to find those logs. Nginx doesn’t seem to have any problems which it sees as problems. Ie. there are no errors to speak of. Other than that I haven’t tried to find logs, as I think it would be in something like “syslog” so would need to have some idea as to what to grep for.

Still none of the follows (from last look, I need to look again) are showing up on my instance.

Again, thank you.

I haven’t seen any change here. I’m probably going to do some digging to see if there is any indication that I can find that indicates what might be failing. I’ve heard twice (or maybe more) that improper IPv6 configuration is often a problem, but while I’ve seen some “improper IPv6 configuration,” where I can tell that’s the case, as of now, I’m not sure it is the case here.

Hmm… Though I’m not entirely convinced. While I am seeing connections working on IPv6 incoming, I do recall that I’ve had issues that for some reason outgoing (in the past) wouldn’t work for syncing the package tree. Certainly something to look into…

If remote messages can’t go out over IPv6, Mastodon won’t be able to fetch the “user representation” for incoming posts from servers that use IPv6, and it won’t be able to display those messages.

the specific log you would be looking at in that instances is journald -u mastodon-sidekiq (when using the systemd services)

Is it possible that removing IPv6 would be preferable to having it? It seems that I wasn’t able to connect out, though I was seeing incoming connections.

I’d rather have the support, but unless someone hasn’t got IPv4 support then that should work, and as has been noted various places, IPv6 is often misconfigured.

From what I can tell, I now have IPv6 properly setup, but I’m not even sure how I would disable it. Especially now that I have confirmed that it works (from what I can tell) from my end.

If things are working, probably no need to change them. But you can also edit gai.conf to prefer ipv4 addresses: