2 months ago
Instance deploys "successfully" but is not actually reachable.
On investigation, discovered the following error in logs: invalid port in resolver "fd12::10" in /config/nginx/resolver.conf:3.
Resolved by the following:
SSH in
cd config/nginx
sed 's/fd12::10/[fd12::10]/' resolver.conf
You're welcome. Please fix this at source, so the next person doesn't have to manually modify a file while the instance is running.
3 Replies
2 months ago
Nginx is super picky with IPv6 in the "resolver" directive it definitely needs those square brackets to parse the address correctly. You need make sure the config template gets updated at the source so nobody else has to sed their way out of it.
2 months ago
Nginx tries to read those colons as ports if they aren't wrapped in brackets [], which is exactly why it was throwing that error.
You need to fix the internal resolver.conf template so it handles IPv6 addresses correctly by default.
In the meantime, so you don't have to manual SSH every time, you can toss that sed command into your start script or Dockerfile as a temporary workaround.
Goodluck with you codebase. Hope this helps
2 months ago
Fixed. Thanks for your feedback!
Status changed to Awaiting User Response Railway • 2 months ago
Status changed to Solved andreidrang • 2 months ago