2 months ago
Hi,
Our Node.js service on Railway is experiencing recurring outbound connection
timeouts to fcm.googleapis.com (Firebase Cloud Messaging). All push
notifications sent through the Firebase Admin SDK are failing.
WHAT WE SEE
The TCP connection to port 443 times out at connect time — no TLS handshake,
no HTTP response. Sample errors from our logs (2026-08-16, UTC):
19:19 connect ETIMEDOUT 216.239.38.57:443
19:24 connect ETIMEDOUT 216.239.36.57:443
19:29 connect ETIMEDOUT 216.239.36.55:443
19:34 connect ETIMEDOUT 216.239.36.55:443
All three addresses are in Google's 216.239.32.0/19 range and all resolve from
fcm.googleapis.com. DNS resolution itself succeeds — the failure is purely at
TCP connect.
Happy to run any diagnostics (traceroute, MTR, curl timing) from inside the
container if that helps narrow it down.
Thanks,
Evren
1 Replies
2 months ago
Your service's egress flow records show zero dropped packets and other outbound connections completing normally, which rules out a Railway-side egress or routing fault. The TCP connect timeouts to fcm.googleapis.com are between your application's client and Google's endpoint, not a platform-level block. The two places to look are the connect/socket timeout configured in the Firebase Admin SDK's HTTP transport (the default in many Node.js HTTP clients is 120 s, and a lower value in a prior environment can surface here) and Google's own quota or rate-limit console for your FCM project, since quota-exceeded responses from Google can surface as unanswered SYNs rather than HTTP 429s.
Status changed to Awaiting User Response Railway • about 2 months ago
a month 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 • about 1 month ago