Not really, but I can empathize with anyone running a relay it can feel like being attacked all the time. Nostr is a client-server architecture where the server has no idea what the clients are doing or when they are releasing or etc. Many times I've found it's just inefficient client side code (including blastr, aggregators, scrapers, and clients with large or temporarily large user base) that can easily slam the relays. So it is a very hard environment to work in being on the relay side. Over time, all the hardening has paid off and I don't notice it much anymore. This all gets wrapped into relay.tools OSS and should help self-hosters survive on the public nostr net should they choose to run it.
Discussion
Clients unintentionally DDoSing relays, as well as writing all sorts of junk into replaceable and addressable events until you inevitably lose data, is basically another Tuesday for anyone running, well, any Nostr software. I don’t know how folks like you, Semisol, Will and Somebody manage this at scale, sir. Thank you for your service.
Just to clarify, though, not everything that happens on Nostr is down to crappy clients. After a while, you get the filters, rate limits, logging, alerts, automatic remediation scripts, dashboards, caching, backup, restart policies, ercsomewhat right; even on personal relays running on a potato.
We still have our fair share of script kiddies and a culture of people with personal vendettas over stuff that only exists in their heads hanging around here. Unfortunately, this happens much more often than we like to admit. All of this "See. I broke your relay/client/whatever. See, see, I was right, I won, I won!" behaviour is just… toddler-level antics by people who are too old for this, should have better things to do, and, most of all, should know better. I'm tired Sir!
yah, its still one of the hardest problems to solve, is how to keep relays online. the cool part is usually they aren't *all down at once if we all keep runnin' them 🐳