the _acme-challenge CNAME had been proxied by Cloudflare, so renewal could never validate

App kastov, org nicolas-615, hostname supo.live.
The apex certificate expired 2026‑09‑25 17:50 UTC (my fault — the _acme-challenge CNAME had been proxied by Cloudflare, so renewal could never validate). I fixed the DNS at 11:50 UTC on 09‑26. www.supo.live issued immediately. The apex has never issued since.
Current state: clientStatus: “Awaiting certificates”, issued: , validationErrors: , rateLimitedUntil: null, re-requesting every ~6 minutes for 12+ hours. checkCertificate returns correct aRecords / aaaaRecords matching the app’s own IPs, isConfigured: true.
What I ruled out:

- Not DNS. The _acme-challenge CNAME is correct at Cloudflare’s authoritative nameservers and the TXT at supo.live.m1j8lxe.flydns.net matches.

  • Not Let’s Encrypt. newOrder for supo.live returns 201, and I completed a DNS‑01 challenge for it on an independent ACME account — LE returned valid at 13:56:24Z. No CAA on the zone.
  • Not the pipeline. certtest.supo.live, created fresh in the same zone and app, issued in 45 seconds.
  • Not the challenge type. I reconfigured the apex to match supoclips.com exactly — the other apex on this app, which renews fine: grey-clouded, no _acme-challenge record, isAcmeAlpnConfigured: true. Still nothing.
  • Not a conflicting app. Org-wide scan; supo.live is only on kastov.
    One detail that may matter: the published DNS‑01 token bdhY5f2rTsCLlApbtUkwK8ZPwPOVxQr5iNY8tUhSNu0 has not changed in 12 hours, across two certs remove / certs add cycles. LE only reuses a token while its authorization is still pending, which suggests the authorization has never been submitted for validation.
    Site is currently served over Cloudflare Flexible as a stopgap, so the origin runs without TLS. Would appreciate someone checking the ACME order state for this hostname.

hi — thanks for the detailed report! we’re looking into this and will keep you updated.

just checking back in to let you know that we we’re still working on this. We’ve pinpointed the problem, though, and are working towards a fix.

@flyio-support Hi, we’re seeing the same issue. App proxy-web-app, hostname proxy.peacockparent.com.

Renewal failed and the cert expired on Oct 6. After we fixed DNS, a new cert was actually issued (Oct 7 ~03:16 UTC, visible in CT logs) and fly certs show reported “Issued”, but the edge served no certificate for the hostname.

We removed and re-added the hostname. Since then the dashboard shows “Issuing…” (fly certs show: “Awaiting certificates”, certificates: []), even though dns_configured and ownership_txt_configured are both true.

Could you include this hostname in the fix? Thanks!

Hi all! Chiming in to say the fix is still in progress. If it doesn’t land this week, cert issuances that are stalled because of this bug should still resolve on their own within 7 days of when they got stuck.

@peacock-parent you should see that hostname’s cert issued on Oct 14 at the latest. I’ll see if there’s any way you can be unblocked sooner.

@supo I checked our logs and your cert was issued on Oct 3 ~13:00 UTC. It looks to me like you’re all set!

We’ll keep working on this fix and update the thread when we have news.