Federation not working anymore after Upgrade to 5.1 (Dockerized)

Hello.

I’m administrating a peertube instance which is federated by several instances. New videos were at leat federated by 2 instances.

Now we switched from manuell installation to dockerized setup and new 5.1 Version.

I can now see, that new videos are not federated by other instances anymore.

I already did a lot of research but i can’t see no errors or problems. Sadly the information about the federation system is low. How can i debug the process of federating new videos?

Regards,
Oliver

Did you restore full database when migrating?

In your instance webadmin page, are they any instance listed as following yours?

Maybe your instance was shutdown too long, and other instances stop federating?
You can try to go to the about page on the 2 other instances, and check if they are still following you.

If you are registered on PeerTube instances , you can check the health of your instance.

Or maybe your instance was spamming and other instances have blocked yours. Are you responding to spam reports?

Hi John.

Thanks for your reply.

Yes, i did a full database backup while migrating and everything went fine. I had every data, as before.
It also was no long shutdown, just a few minutes and dns change.

I can still see the other instances following in admin area and if i visit the other instances they still list my instance on about page. They also are in fact federating the old videos (before the migration).

I would see spam reports, but i didn’t got any.

I think it is something in the process of proposing new videos to the following instances, but there is so little information about that process and how i can debug it.

Any idea?

Regards,
Oliver

Have you checked error logs on the server?

Yes and found nothing suspicious.

Found something in admin-ui « jobs » named « federate-video »:

{
  "videoUUID": "6aa4e141-8c9a-49fc-cc7d-ce7910ed51a6",
  "isNewVideo": false
}

After i edited the video, but still no federation and no further information.
The job itself went successfull.

One thing, that came into my mind is, that i didn’t copy the configuration files from the old server, but instead copied all values from the old admin-ui into the new one.

I found some jobs mentioning something about Signatures. Could it be i accidently deleted the previous signature and federated servers are discarding messages?

Yes it could. But i think signature keys are in the database, not in configuration files (that was the purpose of my first question, if you restored the full database).

Yes, the database was fully restored. But maybe the database model is broken? Is there a repair function for the database?

I’m stuck in this. I already talked to other instance owners who confirmed that readding my instance is possible, but no videos are backed up.

No info in logs, no info in commands view. How to debug this?