Hi team,
Following up on this thread with a formal escalation request — this has been confirmed as a broken edge node issue, not client-side flakiness, and I'd like it routed to your infrastructure/edge network team rather than continuing in community support.
Summary of findings:
- Domain: marsmedia.cc / www.marsmedia.cc
- From my home network, all connections resolve to edge IP 64.29.17.1
openssl s_client -connect www.marsmedia.cc:443against that IP fails with "wrong version number" — the server is not responding with valid TLS at all, not a cert mismatch or expired cert- From a different network, the domain resolves to a different, healthy edge node and works correctly
- DNS resolution is correct and consistent across networks (ruled out)
- CAA records: verified clean, no restrictive policy blocking issuance
- No stray AAAA or leftover _acme-challenge records found
- This points to the failure occurring at the TCP/TLS layer at specific edge nodes, not DNS or domain configuration
Affected ranges observed:
- 64.29.17.0/24
- 216.198.79.0/24
Request: Could someone on the edge network/infrastructure team check the TLS termination health for nodes in these ranges, specifically 64.29.17.1? Given this is affecting real visitor traffic to my production domain, I'd appreciate this being treated as an infrastructure incident rather than a per-account configuration issue.
Happy to provide additional diagnostics (traceroutes, further openssl output, timestamps of failed requests) if useful.
Thanks,
