@Mastodon how is #scaleability handled on #mastodon? I'm concerned about traffic. Do the messages get aggregated before being sent to various sites? Or is it sitting individual messages to every user that is subscribed?

@denebeim @Mastodon uses a message queuing system called sidekiq that handles passing messages around

@denebeim @Mastodon sidekiq is a commonly used message queuing system so you could look up the docs on it, yeah. you can find a link to your sidekiq under administration in the settings page to see the web frontend for it that shows you stats on how many messages it's processing

@anna @Mastodon well looking at it it seems to be like a sidecar, a process that runs doing long time things. I'm not sure that tells me what I was trying to ask though. I guess I should probably look at the source when I have some time.

@denebeim it processes all the messages, as well as doing long running tasks. if you stop it, pretty much everything stops working. not giving sidekiq enough workers is the most common reason you would experience federation delays or other severe issues

@anna and that's what I was thinking of as the bottleneck. If you're sending the same message to multiple people it seems like it would make sense to do it once instead of many times. Maybe I'm just showing my age most of my federation practice was on Usenet. And transfer rates were much lower then.