fix(staging): retry the TCP connectivity probes that expect an open port #137
Loading…
Reference in a new issue
No description provided.
Delete branch "jfraeys/fix/connectivity-probe-retries"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Staging run API 604 (UI #413, the push for #136) failed in the connectivity check on exactly one probe: 'tailnet Postfix relay (ACL tag:web -> services:25) -> got filtered, wanted open'. Everything else passed in that run (hardening validation on both nodes including the journald check, tailnet SSH, Forgejo API, LLDAP, public ports, HTTPS via Traefik).
Nothing merged since the last green run (UI #408) touches Postfix, the firewall or the ACLs (#135 is a merge record; #136 changed only the trivy script, lint and tests). The TCP probe is a single sample, whereas the HTTPS and LLDAP probes right below it retry 6 x 10 s; a container that is still starting leaves its published port dropped, which the probe reports as 'filtered'. This is the likely cause, not a proven one.
🤖 Generated with Claude Code
jfraeys referenced this pull request2026-10-13 19:03:36 -04:00