With the AP-typed match, the station-mode netdev is unmanaged until
hostapd flips it, so any wpa_supplicant/iwd/NetworkManager in the
importing config would grab the radio first and keep resetting it to
managed mode under hostapd. Fail at eval instead of at runtime.
networkd enslaves the radio to br0 once, at link init. On real hardware
(GL-MT6000, mt798x-wmac) that happens while the netdev is still in station
mode, so the kernel rejects the bridge join (IFF_DONT_BRIDGE), the link is
parked in networkd's failed state and never retried: link_reconfigure_impl()
is a no-op while the matching .network file is unchanged, and the udev
'iw set type __ap' hook meant to pre-empt this loses the race. Wireless
clients associate but their DHCP never reaches vlan-lan.
Match on WLANInterfaceType=ap as well as the name. The station-mode netdev
then matches nothing (unmanaged) and the file first matches when hostapd
has switched the radio to AP and raised carrier, so the enslave succeeds
on the first try. Drop the udev hook.
Verified with checks.x86_64-linux.router (hwsim STA lease over the bridge).
New cnx.router.accessPorts option pins a port untagged to one VLAN via
bridge PVID/EgressUntagged; convention is the last copper port as an
always-available mgmt recovery port. gw-cnx-1 port roles: enp1s0 WAN,
enp2s0 trunk, enp3s0 temporary DHCP uplink into the old OPNsense LAN
(back to trunk at cutover), enp4s0 untagged mgmt.
The newedge.house site adopts the 10.1.<vlanId>.0/24 convention: mgmt 10
(the old untagged LAN), lan 20, iot 40, voip 50, dmz 60, unit1-5 110-150.
PPPoE is untagged at this site. Static leases move to a new per-VLAN
dhcp.reservations option rendered into Kea host reservations.
Per-gateway download/upload/ping from the hourly librespeed runs, with
each link's 7-day median overlaid so the WANSpeedDegraded alert
condition is visible at a glance. Label-driven — new sites appear
automatically once scraped.
iperf3 serves throughput tests from every VLAN and the mesh; an hourly
librespeed run feeds speedtest_* metrics through node_exporter's textfile
collector, and vmalert flags download rates below half the link's own
7-day median so ISP degradation surfaces without per-site thresholds.
cnx.router.proxy serves <service>.<site><n>.cnx.network with a real Let's
Encrypt wildcard obtained via a gateway-scoped TSIG key against ns1; Blocky
resolves the names to the router's LAN address, so they exist only
internally. First user: Omada UI on gw-cnx-1 (omada.cnx1.cnx.network).
sshd carries the admin SSH keys and now also CA-signed host certificates
(TOFU-less verification via the shared openssh-ca); the users service
generates per-machine root passwords (share = false, no prompts).
Reusable cnx.router.* module for the Topton 1U boxes replacing OPNsense:
PPPoE WAN (optionally VLAN-tagged, AIS: 10, secret credentials incl.
username), VLAN-filtering bridge, nftables NAT/firewall with MSS clamp,
Kea DHCP with per-VLAN lease time, Blocky DNS, DHCPv6-PD, CrowdSec with
the ZeroTier mesh whitelisted, optional Omada controller, ZFS disk.
Fleet baseline rides along: admins sops group is now derived for every
machine in clan.nix (secrets encrypt to it from the first vars generate)
and time sync is chrony everywhere instead of systemd-timesyncd.
kurogeek
Berwn