I second @vtez44, You would only host Your own private profile data - like password. Whenever You post to an other instance, then your post (data + username) is being shared to that instance - it does not matter if you connect via your own instance or not. There is only little advantage.
DNS-Names are not bound/specific to a protocol/service type. Of course some might fit better for a serivce, e.g. the top level domain ".im" fits "Instant Messaging" (e.g xmpp, matrix ...) services well. But still you could host a Website (https-protocol) on the same address. The list of valid top level domains is not static. new TLDs are being approved constantly. This means e.g. ".world" is a valid Top Level Domain but not many/big services use it for now, so it is rather unknown.
Officially it is the domain of "Mali" https://en.m.wikipedia.org/wiki/.ml Might be also the letters "l" and "m" which are the consonants of lemmy There is no ".lm" top level domain
It would replicate this one specific group/chat only - from the moment of subscription. Depending on the load it can get heavy on your instance.
Matrix has one caveat: it synchronizes every room (group chat) from another instance to your instance fully- which one user subscribed to on your instance. Because of this the instance-systems/servers are under heavy load for private userage (not controllable number of users and chats). Many governmental institutions (controllable number of users and chats) use though, because in case of "disasters/incidents" the data is not lost but saved all over all replicas.
Fediverse is based on a distributed protocoll, where everyone can join. It gives alternatives to most known "social media" service silos like:
- pixelfed : instagram
- mastodon : twitter
- Lemmy : reddit ...
Go ahead and use a search engine ;)
Use a dedicated DNS-service on your local network which has the local IPs in it's DB. Use that DNS-service as your first/primary DNS on your local network (settings).
Could be, that they are out of sync. Maybe overloaded. Not enough capacity/bandwidth?
Hey, what do you mean by "scale horizontally"? There are multiple approaches to tackle this.
- Have multiple nodes/pods for the same instance and run them on a cloud-like service provider
- have RO-instances to handle to read-load
- share/merge bigger communities/subs over multiple instances
- ...
All of these requiere most likely a major rewrite/change of Lemmy server software I guess. They are already addressed as issues/feature requests on github In my opinion the first option would fit the most.
Maybe one bigger instance was not reachable at that moment due to some maintenance?
Hey, why you think it gets overloaded? Is something slower / lagging in synchronization?