According to their documentation, one of the use cases for Relay is to "Improve event response time in regions with low bandwidth or limited connectivity".
We host our Lambda and EC2 resources in AWS us-east-1 region (N. VA). Before going through the setup and increasing complexity of our infrastructure, can anyone speak to the performance improvement? I understand there is more likely to be an improvement if we were in a remote geographic region with slow internet.
Has anyone tried using Relay and seen improvements in geographic regions with good bandwidth and connectivity?
Sentry already hosts Relay in several locations around the world. This blog post goes in more detail: https://blog.sentry.io/2022/03/31/sentry-points-of-presence-how-we-built-a-distributed-ingestion/
A benefit you can have is if you host a Relay instance in your local network, where several of your servers can connect directly to. This way, the Sentry SDK would have minimal latency to send data to Relay (LAN latency vs WAN latency).
For most users, given the PoP Relay deployment, there's not much value in running your own Relay unless you want other benefits like PII scrubbing on-premise (before it hits Sentry's Relays).