Singapore ingress delays: widget times out at 10s before API receipt; matched 499s
scoots78
PROOP

4 days ago

Please investigate possible Singapore edge or ingress delays on 1 October 2026 UTC.

Our booking widget's Vercel functions in Singapore call an API hosted on Railway in Singapore. Diners and staff experienced intermittent loading and booking availability failures across different networks. Some reads hit the application's ten-second fetch deadline and return HTTP 504 PARTNER_API_TIMEOUT. Railway logs show matching HTTP 499 / Node ECONNRESET: aborted for some requests.

Two correlated examples around 09:21 and 09:40 UTC show the API logging receipt around or after the caller's deadline. In one case the API then completed HTTP 200 in about 11 ms; in the other the incoming connection aborted before a parsed body was recorded. Cross-provider clock alignment has not been independently verified.

Resource checks around the failure window showed low API and database CPU usage, stable API memory, and no sampled database pooler waits or deadlocks. The same API process completed other requests quickly during a failed caller's ten-second interval. Subsequent month-calendar reads completed normally. These checks do not prove recovery or exclude brief stalls.

A related public customer report describes similar symptoms:

https://station.railway.com/questions/edge-sin1-tls-handshake-10-28s-timeou-a72db32f

Another commenter there reports Vercel Singapore to Railway Singapore connection timeouts despite a fast API. A common cause has not been confirmed.

Please investigate whether a Singapore edge, TLS or ingress issue affected this route around 09:21-09:42 UTC. We need edge acceptance, TLS, routing, queuing, retry and container dispatch timings, plus any confirmed incident or mitigation guidance.

Production identifiers and detailed traces have been removed from this post. Please provide a secure private support channel for exact request and infrastructure identifiers if needed for tracing. Please do not reproduce earlier identifiers in any public response.

No restart, redeployment, scaling or timeout change has been performed for this investigation. Please propose any production change for our approval.

$20 Bounty

1 Replies

Railway
BOT

4 days ago

We haven't established what's causing this yet. The community can help investigate your setup.

That's exactly what the Railway community is good at, so we'd like to open your thread as a community bounty. Railway pays a bounty to the community member who solves it, and threads like this usually get picked up quickly.

Opening it makes this entire thread public, including everything already posted. Nothing becomes public until you decide. Use the buttons below.

  • Open to the community - Before you click, take a moment to edit or remove anything you'd rather not share. The thread becomes publicly visible right away.
  • Keep it private and close the thread - Nothing becomes public and the thread closes.

Status changed to Awaiting User Response Railway • 4 days ago


Railway
BOT

4 days ago

This thread has been opened as a public bounty so the community can help solve it. The thread and any further activity are now visible to everyone.

Status changed to Open Railway • 4 days ago


Welcome!

Sign in to your Railway account to join the conversation.

Loading...