Subject: All requests to my deployment have a constant ~12s TTFB delay

Project: beck-sinode / Service: sijemaat-system URL: https://sijemaat-system-production-6acf.up.railway.app Region: asia-southeast1-eqsg3a (Singapore) Plan: Hobby

Symptom: Every request to my deployment takes a constant ~12 seconds TTFB, regardless of:

• Path (including 404 paths and /api/v1/health)

• Method (GET, POST, OPTIONS, HEAD)

• HTTP version

• Connection: keep-alive vs close

• Rapid repeated requests (no warmup effect)

What's NOT the cause (already ruled out):

• Cold start — second/third sequential request is also 12s

• App code — Railway's own HTTP Logs show container responds in ~470ms

• DB — even 404 paths that don't touch DB take 12s

• Region — was US East, changed to Singapore, no improvement

• Manual Redeploy — no improvement

• Server bind (0.0.0.0) — no improvement

• Container resources — CPU is ~0%, Memory ~100MB (not saturated)

• User location — measured from Indonesia and from US, both ~12s

Data points:

• Browser DevTools Network tab Timing: "Waiting for server response" = 12.07s

• Container internal handling time per Railway HTTP Logs: ~470ms

• Gap of ~11.6s is between Railway edge and container

• Response headers show x-railway-cdn-edge: fastly/cache-sin-wsat1880044-SIN, x-cache: MISS on every request

Looks like a Fastly origin pull / routing config issue specific to this deployment. Please investigate.