This is the same failure mode as the community post "Custom domain stuck on 404 fallback after EU West outage", but triggered by the June 11 Service Disruption (10:21ΓÇô16:48 UTC).

• Project ID: c2134d7f-ea9c-4253-a3af-c6b0f350fd85
• Service ID: eff157b4-64cf-4a32-a2c5-941f493b101e
• Custom domain ID: 400b9ae5-036e-4902-a77b-e9a7eaea4868
• Domain: umami.luisterlink.nl
• Timeline: Stuck since ~17:00 UTC June 11 (post-incident)

Symptoms:

• HTTPS on the custom domain returns 404 with x-railway-fallback: true ΓÇö the edge routing table never picked the domain up after the incident
• Certificate permanently stuck in "Validating ownership" ΓÇö the Let's Encrypt HTTP challenge hits the fallback 404, so validation can never complete
• The service itself is healthy: umami-production-e104.up.railway.app serves 200
• Your API reports DNS PROPAGATED (CNAME via Cloudflare DNS-only ΓåÆ ajel6mey.up.railway.app); externally verified: no CAA, no DNSSEC

Everything we've already ruled out by experiment:

• 5├ù custom domain delete/recreate (Cloudflare CNAME updated within seconds each time)
• Full service delete/recreate (fresh service ID)
• Region migration US East ΓåÆ EU West (europe-west4-drams3a) ΓÇö the fallback 404 reproduces identically from the europe-west4 edge, so the broken routing entry is global, not region-local
• Re-attaching the domain after the EU move (fresh domain ID born under the EU-homed service)

Could you force an edge routing re-sync / cache flush for this hostname? Nothing client-side remains to try.