Subject: Private network timeout from app services to Redis after May 20 incident

Hi Railway team,

We’re seeing a persistent private networking issue in our production Railway project after the May 19/20 incident.

Project: cookmark-backend

Environment: production

Region/edge seen: europe-west4-drams3a

Affected services:

API service

Worker service

Redis service/database via private DNS: redis.railway.internal:6379

Current failure:

redis-ping-failed

dial tcp 10.218.190.169:6379: i/o timeout

Example timestamp:

2026-05-20T13:14:55Z

Public API requests return Railway edge 502 because the API process cannot reach Redis during startup:

Application failed to respond

What we’ve verified:

API and worker both use Railway private DNS for Redis: redis.railway.internal:6379.

The private DNS resolves to 10.218.190.169:6379.

TCP connections to that private Redis IP time out.

Redis appears healthy/online from the Railway dashboard.

API, worker, and Redis are in the same region, so this does not look like a cross-region routing issue.

This started after the May 19/20 Railway incident and a container restart.

Redeploying/restarting API and worker did not restore connectivity.

Can you please investigate private network routing/connectivity between the API/worker services and Redis? This looks infrastructure-side rather than an application-level Redis auth/config issue, since the failure is a TCP timeout before Redis auth/protocol.