Proxy cannot route HTTPS (443) despite passing health checks - no healthy instances found

[Seerror]

Hi Fly.io community
I’m facing an HTTPS routing issue on Fly.io and could use some guidance.

App: seerror-backend
Region: bom (Mumbai)

• Machine status: started

• Health checks: passing

• /health endpoint returns 200 OK every 15s

• App binds correctly to 0.0.0.0:8000 using PORT env var

• No errors in app logs

Issue:
Fly proxy fails to route HTTPS with the error:
no known healthy instances found for route tcp/443

Even though health checks are passing, traffic on port 443 doesn’t route.

Has anyone faced this before? Anything specific to check around services, ports, or proxy/TLS config?

Thanks in advance

Seerror

---

[halfer]

In general, it is ideal if questions can be accompanied by technical details, without needing for an an engineer to request them. Would you add to this post your fly.toml and maybe also your Dockerfile? I’d also suggest adding your deploy logs. Please use Markdown code formatting for all three.

[halfer]

There are some things you can do to check things for yourself:

1. Run your app locally, using Docker, and see if it has a long start-up time, or fails to boot;
2. Shell into your app from your local console and use cURL or wget to fetch from port 8000;
3. Launch your app in another region, to see if Mumbai might be experiencing problems.

If you can do any or all of these, and report the results here if the issue is still not fixed, readers will have plenty of material with which to formulate new advice.

[pavel]

Hey @Seerror

I’ve just tried sending a request to your app and it responded just fine. Are you still experiencing the issue?

One thing I noticed - it takes your app roughly a minute to boot and start passing the health checks, so the first request will likely fail, as you have only one machine and proxy won’t be retrying for so long. Once the app boots and passes the health checks, it seems to be responding just fine.