Hello,
When I tried making an instance on AWS and connected it all correctly, my instance is timing out, not allowing me to connect to it.
The website is https://test.darkpixlz.tk, and the server claims it’s running.
AI-generated summary
The user darkpixlz is experiencing a connection timeout when trying to access their instance on AWS, despite the server appearing to be running. They are using Cloudflare to route traffic via a domain, with an A record pointing to the server’s public IP. Stephen and Falco ask whether the EC2 security group allows inbound traffic on ports 80 and 443, which darkpixlz confirms they have configured. RGJ suggests testing with Cloudflare in DNS-only mode, but darkpixlz reports the same timeout. The issue is likely related to Cloudflare’s proxy settings, as Stephen notes that the orange cloud (proxied mode) prevents Cloudflare from reaching the server. darkpixlz resolves the issue by creating a new instance and using HTTP, though Stephen suggests this may not be necessary and could be due to a Let’s Encrypt cooldown if the instance was rebuilt multiple times.
What did you do to make the server available externally? Did discourse-setup do the connectivity check, or did you bypass it?
I’m running the public IP through Cloudflare to access through the domain, and yes it passed.
So you set up specific access rules for 80 and 443?
So what happens if you switch it to “DNS only”? Does it start working then?
When I was doing that I was getting my browser’s “Connection Timed Out” screen.
Have you configured the EC2 instance security group in order to allow ingress traffic on the ports 80/443?
That just means Cloudflare can’t contact an upstream server. Turn off the orange cloud.
I was able to get it working, I just had to make another instance, then use HTTP.
That shouldn’t be necessary.
Did you try to build the instance more than once prior to fixing the access rules? If so it’s possible Let’s Encrypt put your hostname on a seven day cooldown.