@i honestly I canceled the service by accident thinking it was another relay for something else and i lost the database so I never considered setting it up again. never had an issue with bloat, it was running on a like 25$/year vps that had like 80G SSD and it seemed fine. only issue i ever had was before they made attempts to thread it, it would lock up any time mastodon instances would disappear or send mass delete activities but the threading fixed that
people are still relying on relays? they still use em?
@graf i'm the retard with ~2k relays subscriptions instead of follow botting federation on a single user instance, the bloat is on the receiving instance since Announce activities pile up, running the relay itself is a matter of efficiently broadcasting the ap id's
@graf any chance you ever get a relay working again?
you can run https://256.lt/mrf/relay_optimization.ex if db bloat is a concern and running on develop should help with past issues of things getting stuck when instances are down temporarily
@i Just got in to bed but I bookmarked this. I'll set it up again tomorrow if you think people will use it. I was thinking about it a couple weeks ago but thought it might be a lost cause but I will look into it further tomorrow. Will be the same address (relay.fedi.agency)
@i I installed an OS on the server I'm going to use for the relay I, I stalled at picking one to host because they both have flaws. I would have had it done yesterday but we had broken federation for a few hours due to a shitty Erlang deployment (Gentoo removed 25.x). Will look at this shortly again