Reddit chose a very bad time to mess up, right when we’re in the middle of coding some of our biggest performance improvements. They timed this perfectly to make things as stressful as possible for nutomic and I lol.
I’ve been busy too recently and I just noticed yesterday the influx of new users… So I’ve updated and upgraded the server a bit. Hopefully I can keep my instance running smoothly and contribute a little bit that way!
The ones that really want to get in, would wait a long time i guess, like i certainly would! The only reason i am not here months ago, is because i did not know this place existed
Depending on how bad the perf is, it may be worth it throwing the instance on a real big machine to take the load to buy time for the optimization work without turning many users away. Contributions should increase with the flood of users. If a big machine could mask the problems of course. Maybe they’re too bad to mask. You probably know.
Ya… our server upgrade was a short-term fix, but our code efforts in the past week, especially with regard to the SQL and websockets, should help things.
Has there been any work / planning around an ability for third parties to setup broadcast servers to help distribute the load? I know that kind of thing is extremely difficult but maybe it would be a reasonable effort for handling read only requests. I haven’t had a chance to look into lemmy’s architecture yet so not sure if this is a valid question or not :)
It would make a great stress test though when suddenly millions join Lemmy :)
Reddit chose a very bad time to mess up, right when we’re in the middle of coding some of our biggest performance improvements. They timed this perfectly to make things as stressful as possible for nutomic and I lol.
Great work and good luck!!!
I’ve been busy too recently and I just noticed yesterday the influx of new users… So I’ve updated and upgraded the server a bit. Hopefully I can keep my instance running smoothly and contribute a little bit that way!
Well…I hear good things about Ritalin. Good luck, and godspeed lol
I am so not jealous. Just know that we appreciate all your hard work.
The ones that really want to get in, would wait a long time i guess, like i certainly would! The only reason i am not here months ago, is because i did not know this place existed
I just discovered kbin which looks very nice. Do you plan to share the like/dislike feature with them so that accounts can be cross-used?
You might need to open up an issue on the github if likes aren’t federating with it.
Depending on how bad the perf is, it may be worth it throwing the instance on a real big machine to take the load to buy time for the optimization work without turning many users away. Contributions should increase with the flood of users. If a big machine could mask the problems of course. Maybe they’re too bad to mask. You probably know.
Ya… our server upgrade was a short-term fix, but our code efforts in the past week, especially with regard to the SQL and websockets, should help things.
Your efforts are appreciated!
❤
Has there been any work / planning around an ability for third parties to setup broadcast servers to help distribute the load? I know that kind of thing is extremely difficult but maybe it would be a reasonable effort for handling read only requests. I haven’t had a chance to look into lemmy’s architecture yet so not sure if this is a valid question or not :)
Not that I know of.
No stress test needed, every admin has already said they know it’s going down