Account: [email protected] Project: untitled-1 (id 69e1fadde3efc3fd35585784) Service: pencil (id 69e70dd6642e287152775aa1) Custom domain: pencilnow.app Auto-generated domain: pencil-uat.zeabur.app (working normally)
Verified items:
- DNS resolution is correct: pencilnow.app → 52.223.32.133 (Zeabur hkg1 edge)
- Cloudflare DNS: CNAME @ → cname.zeabur.app, proxy set to DNS only (grey cloud)
- TLS certificate issued: CN=pencilnow.app, ZeroSSL DV, valid until 2026-07-31
- Service deployment 69f5b5f67abecea3dc19baf7 is in RUNNING state
- pencil-uat.zeabur.app returns 200, env=production is correct
Issues:
curl https://pencilnow.app/api/healthreturns HTTP/2 404, but the header containsx-zeabur-request-id- This indicates the request reached the Zeabur edge, but the backend host→service routing table has not mapped pencilnow.app to the service.
- Domain status has been stuck in PROVISIONING since 2026-05-02 08:04 UTC.
Attempts made:
- Deleted and recreated multiple times via CLI / Dashboard
- Service restart
- Service redeploy (latest image is healthy)
- Verified DNS via multiple resolvers
Additional info: Earlier, two consecutive redeploys (69f5b2cc / 69f5b376) triggered an ImagePullBackOff, likely due to an image registry GC race condition (rolling pod hadn't pulled the new image before the old one was GC'd). The latest redeploy 69f5b5f6 is healthy; this is likely unrelated to the routing issue but provided as context.
Could you please manually trigger the routing for pencilnow.app, or help investigate why the provisioning is stuck? Thanks!