Skip to content
Security questions: Edge-to-container network isolation and trusted proxy headers
joba19
OP

14 days ago

Hello Railway team,

I'm preparing to launch a small business website hosted on Railway and would appreciate clarification about the security of Railway's networking infrastructure.

I have a few general questions:

  1. When Railway terminates HTTPS at its public edge and forwards HTTP traffic to an application container, is that internal connection isolated from other customers' workloads? Is it encrypted or otherwise protected against interception?

  2. Does Railway guarantee that only its trusted ingress infrastructure can establish connections to application containers using source addresses in the 100.00.0.0/10 range?

  3. Does Railway's ingress guarantee that client-supplied forwarding headers, such as X-Forwarded-For, X-Real-IP, and X-Forwarded-Proto, are overwritten or sanitized before reaching the application?

  4. Are these behaviors officially documented security guarantees, or implementation details that could change?

I'm particularly interested in official documentation or recommended security practices for applications handling employee authentication.

Thank you for your help!

Solved

1 Reply

Railway
BOT

14 days ago

Traffic from our edge to your container runs over an encrypted WireGuard tunnel, the same transport our private networking uses, so your app seeing plain http:// does not mean that hop is unencrypted. Private networks are isolated between projects and environments, but services within the same project environment share one trust boundary and can reach your app's listening port directly over the private network.

There is no guarantee about the 100.64.0.0/10 range. The proxy address your container sees as its socket peer is not a published or stable range and can change, so it should not be used as a trusted-proxy allowlist or as proof that a request came through the edge.

For requests through our public HTTP edge (custom domains and .up.railway.app), we overwrite X-Real-IP, X-Forwarded-For, X-Forwarded-Proto and X-Forwarded-Host, replacing any client-supplied copies rather than appending to them. Requests arriving over private networking or a TCP proxy do not pass through that sanitization, so another caller on those paths can send its own forwarding headers.

The header meanings are described in our public networking specifications, which note that this information is subject to change, and private network isolation is covered in how private networking works. For employee authentication, use X-Real-IP for client-IP logic such as rate limiting, avoid trusting forwarding headers from every peer if other services in the environment could reach the app, and keep authentication enforced in your application rather than relying on network source addresses.


Status changed to Awaiting User Response Railway • 14 days ago


Railway
BOT

7 days ago

This thread has been marked as solved automatically due to a lack of recent activity. Please re-open this thread or create a new one if you require further assistance. Thank you!

Status changed to Solved Railway • 7 days ago


Welcome!

Sign in to your Railway account to join the conversation.

Loading...