Hi all, just deployed a self-hosted instance and we're running into a deployment issue. As you can see we are getting a "no connection to server" log twice. If we inspect the console, we're seeing errors loading resources from the api route. I've validated that we have all of the cloudformation resources we need that this video walks through: I've also attached the records in our hosted zone. Looks like the api route record is there. We have 8 records. Would love to get some context on why we might be seeing this issue. Thanks!

looks like the server is ok https://api.kipshealth.com/healthcheck - I suspect CORS. What do your settings for allowedOrigins look like? CORS: https://www.medplum.com/docs/self-hosting/setting-up-cors Server Configuration: https://www.medplum.com/docs/self-hosting/config-settings

Really - maybe locally something is cached for me. This is what I get on https://api.kipshealth.com/healthcheck

what do you see in incognito? or curl?

Update to everyone: It indeed was a DNS issue. Our configuration is slightly different than most. We have Cloudflare managing DNS for a Namecheap domain. We also had our root domain pointing to our Vercel app with subdomains managed through Webflow. Just a lot of DNS records in different places. How we fixed it: Made sure we got rid of any app nameserver records in Cloudflare that weren't specifically that Route 53 ones. This meant cleaning up the Vercel and Webflow recorrds. We also made sure our Namecheap domain had all the nameservers from Cloudflare and Route 53 on there. Then for some reason my local machine still gave me an error on DNS even after a couple hours. I ran

sudo killall -HUP mDNSResponder

just in case to clean my DNS cache.