MrKaplan

Hello, I'm part of the Lemmy.World infra team. I'd be happy to help finding the root cause of this issue, but it's currently not clear what information we should be...

> Does the bug get triggered by running multiple Lemmy processes with `--disable-activity-sending`, `--disable-http-server` etc? We've previously been using multiple federation containers and now we are using a single federation...

We also noticed some very strange behavior where it seems that our outbound federation progress for some instances seems to be jumping between two very different states: federation log ```...

it doesn't seem to affect all instances, and it is probably only noticeable for instances with some lag already, but this graph shows instances that are reducing lag without jumping...

From the most recent occurrence for us, we had a short DB outage, which resulted in messages like ``` task exited, restarting: Err(Error occurred while creating a new object: the...

For the case of a user on a 0.19.5 instance unfeaturing a post in a 0.19.3 community, which was reported [here](https://discuss.tchncs.de/post/18316214), I believe this is the associated log from lemmy:...

it seems likely that the way that these entries were created in the db is no longer possible in 0.19.5, as a lot of 0.19.5 instances don't seem to have...

I talked to jwr1 on Matrix and there were some config issues a while back related to HTTP/HTTPS and certificates, and during the troubleshooting/mitigation attempts there may have been responses...

The prime example would be lemmy referencing profiles of remote users as name@domain. IDK is AP allows non-unique `name`s on the same domain, if it does then it seems to...

my development target is currently 0.19, as that is the version we run on lemmy.world. when the 0.19 PR is acceptable overall I can create an additional PR against main,...