#[1] ’s nostream postgres relay seems to be performing much better than nostr-rs-relay under a large amount of concurrency. I am going to switch the damus relay to that. It will also allow me to scale up even further in the future using read replicas.