8081 is the alt-HTTP port — one of the crowded defaults every proxy/dev server
grabs — so it's collision-prone on a shared host. Move to 22672, a registered-
range port in the sparse 20000s band, derived deterministically from the
service name and recorded in architecture port-allocations.md.
Updated everywhere the number appears: api bind config + default, firewalld
service XML, nginx upstream, vite dev proxy, deploy workflow API_PORT,
infra-setup (SELinux semanage label), and docs.
Operator note: re-run script/infra-setup.sh on slartibartfast to relabel the
SELinux port (22672) and ship the updated firewalld XML before the next deploy
restarts the api on the new port.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_016fKZzDpvjiJ9eYbPGgJvUP
infra-setup only handled the internal (newsfeed.internal) cert and vhost and
just printed a generic "set it up yourself" note for the WAN name. Now it also,
on the web host:
- issues the Let's Encrypt cert for rob.fyi idempotently (certbot, Cloudflare
DNS-01, ECDSA, --keep-until-expiring), skipping cleanly if the CF token or
certbot is absent;
- installs the certbot renew deploy-hook (reload nginx);
- installs the public vhost, gated on the cert existing so a missing cert can't
break `nginx -t` for all of nginx (external-tls.md §4).
PUBLIC_FQDN/CERT_EMAIL/CF_CREDS are infra-truth vars at the top; set PUBLIC_FQDN
empty to skip. Cloudflare A record + OPNsense :443 forward remain manual and are
called out in the closing output.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_016fKZzDpvjiJ9eYbPGgJvUP
Reverts the move to `infra`. Verified on a fedora-43 runner that the ssh
reachability check ran fine (it added the host key and connected) — the
earlier deploy failure was the unprovisioned remote sudo step, not a missing
ssh client. Every runner has ssh: `git` (installed for checkout) pulls
`git-core`, which hard-requires `openssh-clients`. So fedora-43 (ssh + rsync +
curl) is the right, lighter runner for these deploy jobs.
See architecture gitea-runners.md.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_016fKZzDpvjiJ9eYbPGgJvUP
Run #2 exposed two runner-mismatch bugs (now documented in architecture
gitea-runners.md):
- build-web failed with `corepack: command not found` — the fedora runners
ship pnpm preinstalled via npm, not corepack. Call pnpm directly.
- deploy jobs need an ssh client, which the base fedora images lack (they have
rsync but not openssh-clients). Move deploy-api/deploy-web to `runs-on: infra`.
build-api already passes on `runs-on: rust`.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_016fKZzDpvjiJ9eYbPGgJvUP
The first run failed with `cargo: command not found` — `runs-on: fedora-43`
is the generic (node) runner, not the Rust toolchain. Split the build:
Rust on `runs-on: rust`, the SPA on `fedora-43` (matching the fleet's
existing conventions, cf. cichlid/helexa workflows).
Also drop the musl static-cross build: the entire fleet is Fedora 43, so
native binaries built on the `rust` runner run on the targets without glibc
skew (deployment-gitea-actions.md §6 permits building on a runner no newer
than the oldest target). Removes the unverifiable musl-toolchain dependency.
Pin pnpm via packageManager so corepack resolves it deterministically.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_016fKZzDpvjiJ9eYbPGgJvUP
