an hour ago
Account: signed in as the project owner (JoeNayer). Project: hospitable-imagination (4076af4b-a9aa-4a75-b2c8-88e551189ab8). Service: shopdock (53f73c78-4842-4b36-b706-eae992d86ec7), environment production. Current instance: 9f289f10 (deployment of 2026-10-23 09:23 UTC).
Problem: the service's web Console has not connected at any point today. Every attempt ends at "Disconnected" with "WebSocket connection failed" in the console body; no shell prompt has ever appeared. This persisted through the platform incident QH8L8B3E ("Railway Platform Outage") AND after it: the incident was marked Resolved at 2026-10-23T22:06:13Z, the outage banner has disappeared from the dashboard, and the console still fails identically.
Attempts (all while signed in; dashboard pages otherwise load normally):
- 2026-10-23 ~09:00–09:39 UTC — Console stuck on "Connecting..." (embedded view, popout, reloads, both deployment instances).
- 2026-10-23 12:09 UTC — "Disconnected" / "WebSocket connection failed", full 4-minute watch.
- 2026-10-23 19:30 UTC — same end state, 3+ minute watch (incident in Monitoring at the time).
- 2026-10-23 22:16–22:21 UTC — same end state, full 3-minute watch, AFTER the incident resolved and the banner cleared.
- A connection attempt via the Railway CLI (railway ssh, same account) on 2026-10-23 at 22:26–22:31 UTC failed with: the CLI login could not complete — its device-auth request to backboard.railway.com timed out repeatedly from the operator's machine (the same machine reaches backboard via curl in under a second), so railway ssh was never reached.
Meanwhile the service itself is healthy and serving production traffic throughout (public health endpoint 200; current deployment active since 09:23 UTC). The failure appears specific to the console/websocket path for this service/instance. Please investigate.
A follow-up attempt via the Railway CLI from a second machine on the owner's network (railway ssh, same account) on 2026-10-23 at 22:47 UTC CONNECTED and landed a shell in the production container — the fault is specific to the web console's browser-websocket path.
1 Replies
Status changed to Awaiting Railway Response Railway • about 1 hour ago
20 minutes ago
Console and SSH failing on that service isn't caused by a static IP setting or anything in your configuration, so there's nothing you need to change. We've passed this to our engineering team with the details you sent. Your running services aren't affected.
Status changed to Awaiting User Response Railway • 20 minutes ago