Internal Network Connectivity Issues & Dashboard Unresponsive
imadraude
PROOP

2 months ago

Hello Railway Support Team,

I am experiencing severe network connectivity issues with my project. It appears to be a network partition or a stuck networking stack on the node hosting my services.

Project ID:32b8d984-049f-48ab-a6f6-a920a59e3598

The Symptoms:

  1. Dashboard Issue: The PostgreSQL service page in the Railway Dashboard is stuck on "Attempting to connect..." indefinitely. It does not show metrics or status, even though the deployment is marked as "Online".

  2. Internal Network Instability: My Python services frequently fail with httpx.ConnectTimeout when trying to reach external APIs (like Telegram) or when resolving internal DNS, despite no code changes.

Troubleshooting I have performed:

  1. I verified that the PostgreSQL database is actually running and accessible externally. I successfully connected to it via the TCP Proxy (Test-NetConnection to shuttle.proxy.rlwy.net succeeded).

  2. I have attempted to Restart and Redeploy both the database and the services multiple times, but the networking issues persist.

Could you please investigate the internal networking for this project or reset the networking stack?

Thank you.

$10 Bounty

1 Replies

ray-chen
EMPLOYEE

2 months ago

This thread has been marked as public for community involvement, as it does not contain any sensitive or personal information. Any further activity in this thread will be visible to everyone.

Status changed to Open ray-chen • 2 months ago

yusufmo1
PRO

2 months ago

This looks like a IPv6/private networking issue. Railway's internal network uses IPv6 by default, and httpx in Python defaults to IPv4 resolution which causes ConnectTimeout errors.

For your Python services, try forcing IPv6:

transport = httpx.HTTPTransport(local_address="::")
client = httpx.Client(transport=transport)

the main thing is :: which means you will bind to ipv6

Loading...