7.4 KiB
homey
Architecture
NixOS homelab on Raspberry Pi 4 (8 GB). Domain: zakobar.com. Static IP: 192.168.1.100. nixpkgs pin: nixos-25.05.
Container runtime: Podman via virtualisation.oci-containers. All containers join the homey podman network (created by podman-homey-network.service in common.nix). Inter-container DNS via container names; Caddy proxies via 127.0.0.1:<port>.
Reverse proxy: Caddy with Cloudflare DNS-01 wildcard cert for *.zakobar.com. Built via pkgs.caddy.withPlugins with github.com/caddy-dns/cloudflare@v0.2.4 (resolved hash: sha256-pRrLBlYRaAyMYwPXeTy4WqWNRu/L9K6Mn2src11dGh8=). Each vhost generates a pair: HTTPS vhost + http:// vhost for cloudflared loopback. Authelia forward_auth uses /api/authz/forward-auth (v4.38+ endpoint, not legacy).
Auth stack: OpenLDAP ← Authelia (TOTP 2FA). Authelia config is rendered entirely at build time from Nix strings. A NIXOS_CONFIG_HASH env var on the container forces a restart whenever the config changes (bind-mounts resolve symlinks at start, so the running container would otherwise keep the old config).
Monitoring: Prometheus (9090) + node_exporter (9100) + systemd_exporter (9558) + Grafana (3002). Node Exporter Full dashboard pre-provisioned (fetchurl hash resolved: sha256-1DE1aaanRHHeCOMWDGdOS1wBXxOF84UXAjJzT5Ek6mM, ID 1860 rev 37). Grafana proxy auth: Authelia → Remote-User header → Caddy maps to X-WEBAUTH-USER → Grafana auto-signs-in. All reaching Grafana are confirmed admins (Authelia enforces two_factor + admins group).
Uptime Kuma sync: a Python oneshot service runs after container start. Hash-gated — only re-syncs when the monitor JSON changes. Supports keyword (keyword monitor type) and maxretries fields beyond the basic name/url/interval.
Attic (self-hosted Nix binary cache): cache name main, public key main:9SZt/6plBU7jjQzz90J7O011I13hmJvOMYouxNqExNQ=, endpoint https://attic.zakobar.com/main. Setup completed 2026-05-30. NAR content not backed up — reproducible from source. Tokens are stateless JWTs; regenerate with same atticadm command if lost.
Eurovision Vote: Django app sourced from external flake github:anerisgreat/eurovote, wrapped by modules/services/eurovote.nix. Uses DynamicUser + StateDirectory so systemd owns /var/lib/eurovote/; no tmpfiles entry needed.
Backup: Restic daily at 03:00 to S3 (Backblaze B2, bucket zakobar-home-backup). Pre-hook: Nextcloud maintenance mode on + pg_dump. Post-hook: maintenance mode off. Manual offload: restic copy to local disk. NAR content and media excluded.
Reliability hardening in hosts/pi-main/default.nix:
- Hardware watchdog:
bcm2835_wdtkernel module, systemd watchdog runtimeTime=300s / rebootTime=360s - WiFi power save disabled: brcmfmac driver drops connections under low traffic; disabled via
iwon interface up - Network watchdog: timer every 2 min (starts 5 min after boot), pings gateway, restarts wpa_supplicant, reboots if still dead after 30s
- zramSwap: zstd, 25% RAM (~2 GB) — breathing room for PHP upload spikes
- Nix build-dir:
/mnt/data/nix-build— avoids small tmpfs filling during large builds
Bootstrap: pi-main-bootstrap config builds an SD image (sd-image-aarch64.nix) for first flash.
Conventions
homeyConfig specialArgs (passed to every module): domain, organization, timezone. Never hardcode domain strings.
users.mutableUsers = false — all user config must be declared in Nix.
Secret injection pattern: LoadCredential stages sops-decrypted file into $CREDENTIALS_DIRECTORY before Exec*; shell script reads and exports. Ephemeral env files written to /run/ and cleaned up in ExecStopPost / postStop.
Authelia accessControlRules: sorted by priority at build time (lower = first). Authelia stops at first match, so more-specific rules must have lower priority. Ranges: 0=bypass, 10–19=blanket bypass, 20–49=admin two_factor+deny pairs, 50–64=one_factor open, 65–79=per-path (resources + subject combos).
accessControlRules option is declared unconditionally (not inside mkIf cfg.enable) so any module can contribute rules even when Authelia is disabled. Same pattern for homey.monitoring.monitors.
Attic token generation: run atticadm make-token inside the container. Tokens are stateless; losing one means just regenerating it with the same command.
DynamicUser services (Eurovision Vote): secrets must be mode 0444 (not 0400) because DynamicUser gets a random UID that cannot be pre-assigned as owner.
Caddy Cloudflare plugin secrets: uses LoadCredential + ExecStart override (clears list with empty string first, then sets the real start command) to export CLOUDFLARE_API_TOKEN before exec-ing caddy.
Gotchas
hdparm APM udev rule was removed — USB-SATA bridges often don't support APM commands and hdparm hangs indefinitely, causing boot-time crashes. hdparm is still available as a package for manual use.
storage.nix config is gated on lib.mkIf (cfg.device ! "")= — if homey.storage.device is empty string, the mount and all tmpfiles rules are skipped. Useful during initial setup.
Grafana login form disabled (disable_login_form = true) — recovery requires re-enabling it in the Nix config. All proxy-auth users are auto-assigned Admin role (safe because Authelia already restricts to admins group).
Nextcloud preview generation: a separate oneshot service nextcloud-generate-previews (declared in hosts/pi-main/default.nix) fills missing thumbnails after first start. Must be triggered manually or via timer.
Authelia config bind-mount gotcha: NixOS resolves the symlink to the nix store path at container start. Without NIXOS_CONFIG_HASH env var, a config change would not take effect until manual container restart.
WiFi network name: Zakobar. sops secret key: wifi/psk. The secret file must contain exactly one line: wifi_psk=<password>. The ext:wifi_psk format is wpa_supplicant's literal substitution syntax, not an env var.
Attic: writing ephemeral TOML config (not a real file in the store) via ExecStartPre shell script that writes to /run/attic-config.toml. JWT secret interpolated into the TOML at runtime.
Key Files
flake.nix— module list,mkHostbuilder,homeyConfigspecialArgs,rpi4Headlesshardware snippethosts/pi-main/default.nix— enabled services, static IP, WiFi, reliability hardening, Attic substituter confighosts/pi-main-bootstrap/default.nix— SD card bootstrap imagemodules/caddy.nix—virtualHostsoption, dual vhost generation, Authelia forward_auth snippetmodules/services/authelia.nix— access control rule rendering,accessControlRulesoption (unconditional)modules/services/uptime-kuma.nix—homey.monitoring.monitorsoption (unconditional), sync scriptmodules/services/attic.nix— Nix binary cache, JWT token config, netrc injection for Nix daemonmodules/services/attic-setup.md— post-deploy steps, token commands, client config, setup historymodules/services/eurovote.nix— DynamicUser wrapper for external flake modulemodules/monitoring.nix— Prometheus + Grafana, proxy auth wiring, Node Exporter Full dashboardmodules/common.nix— Nix settings, podman network creation, sops global configmodules/storage.nix— external HD mount,extraDirsoptionmodules/backup.nix— Restic, pre/post hooks,extraPathsoption