agent: dial the manager by a fleet DNS name (manager.fleet.lab)

The agent's baked default pointed at the PUBLIC admin-UI name
(diskless-server.bitasylum.org), which resolves to the serving host's outbound IP
— unreachable from the fleet LAN. There's no split-horizon to "automate": the
manager agent plane is just another *.fleet.lab, served by the fleet dnsmasq like
boot/nvme/audit.

- serving/.env.example: MGMT_HOST=manager.fleet.lab.
- serving/docker-compose.yml: dnsmasq --address=/${MGMT_HOST}/${SERVER_IP}.
- files/agent/agent.env + main.go default: FLEET_AGENT_MANAGER=manager.fleet.lab:9600.
- server-pki.sh: AGENT_SANS default -> manager.fleet.lab (cert SAN matches the
  baked name, so no FLEET_AGENT_SERVER_NAME override).
- docs/TODO #14 gotcha #4 updated: fleet name, not the IP+server-name workaround.

The public diskless-server.bitasylum.org stays the ADMIN-UI name (Traefik/:443).

docs/TODO.md

@@ -1447,12 +1447,17 @@ the design forks resolved:
   3. **Distroless perms.** The nonroot manager (uid 65532) can't read `pki/agent/*`
      if created `0600` by the host user → `open …/server.crt: permission denied`.
      `chown 65532` the agent PKI (as with the fleet-data volume).
-  4. **Wrong manager IP / split-horizon DNS.** `diskless-server.bitasylum.org`
+  4. **Manager wasn't a fleet name.** The agent's baked default pointed at the
-     resolved to the serving host's OUTBOUND IP (`no route to host` from the fleet
+     PUBLIC admin-UI name (`diskless-server.bitasylum.org`), which resolves to the
-     LAN). Bench workaround: `FLEET_AGENT_MANAGER=<fleet-lan-ip>:9600` +
+     serving host's OUTBOUND IP → `no route to host` from the fleet LAN. FIX: the
-     `FLEET_AGENT_SERVER_NAME=<dns-name>` (dial the IP, verify the cert's SAN).
+     agent plane is now just another `*.fleet.lab` — `manager.fleet.lab` (`MGMT_HOST`
-     PROD FIX (not code): fleet dnsmasq `--address=/<name>/<fleet-lan-ip>` so the
+     in serving/.env), served by the fleet dnsmasq to `SERVER_IP` exactly like
-     baked default resolves on the host VLAN.
+     boot/nvme/audit (`--address=/${MGMT_HOST}/${SERVER_IP}`), baked as
+     `FLEET_AGENT_MANAGER` and set as the agent server cert SAN (`AGENT_SANS`). No
+     per-host override, no server-name hack. (The public `diskless-server…` name
+     stays the ADMIN-UI name on Traefik/:443 — a separate plane.) The transient
+     bench workaround was `FLEET_AGENT_MANAGER=<ip> + FLEET_AGENT_SERVER_NAME=<name>`;
+     no longer needed.
   5. **Host not in the manager's inventory** → `PermissionDenied` (surfaced as a
      bare `EOF` on the agent before the fix). A host enrolled OUT OF BAND (bench
      broker / migrated fleet) has no inventory row. FIX: `fleet-manager

files/agent/agent.env

@@ -6,10 +6,13 @@
 # environment by replacing this file (it is part of the immutable image, so a
 # change means a rebuild + re-sign, like any other baked config).
 
-# The fleet-manager agent plane (gRPC/mTLS). A DNS name, NOT a per-host value.
+# The fleet-manager agent plane (gRPC/mTLS). A fleet DNS name (NOT per-host),
-# Reached from the authenticated host VLAN via the OPNsense L3 hop to the mgmt
+# resolved by the serving dnsmasq like boot/nvme/audit — MGMT_HOST in serving/.env
-# plane (not the boot nginx). Set to your manager's real name:port.
+# points it at the manager's fleet-LAN address (this serving host now; a routable
-FLEET_AGENT_MANAGER=diskless-server.bitasylum.org:9600
+# mgmt address once the manager moves behind OPNsense L3 / k8s). Must match the
+# agent server cert SAN (server-pki.sh AGENT_SANS). This is the AGENT plane, NOT
+# the public admin-UI name (that's Traefik/Let's Encrypt on :443).
+FLEET_AGENT_MANAGER=manager.fleet.lab:9600
 
 # Fleet CA that verifies the manager's server cert. This is the SAME step-ca root
 # that signs the host TPM certs + the RADIUS server cert, so we reuse the in-place

files/agent/main.go

@@ -198,7 +198,7 @@ func logDown(down *fleetpb.ManagerDown) {
 // ---- config ----------------------------------------------------------------
 
 func loadConfig() config {
-	manager := env("FLEET_AGENT_MANAGER", "diskless-server.bitasylum.org:9600")
+	manager := env("FLEET_AGENT_MANAGER", "manager.fleet.lab:9600")
 	cfg := config{
 		manager:    manager,
 		serverName: env("FLEET_AGENT_SERVER_NAME", hostOf(manager)),

serving/.env.example

@@ -72,6 +72,13 @@ NVME_HOST=nvme.fleet.lab
 # Remote audit collector name. Host-generic — bake into the image's rsyslog
 # forward rule so every host offloads auditd here; resolved to this serving host.
 AUDIT_HOST=audit.fleet.lab
+# Fleet-manager AGENT plane name. Baked into the image (files/agent/agent.env
+# FLEET_AGENT_MANAGER) so every enrolled host dials the control plane here;
+# resolved per-environment to the manager's fleet-LAN address (this serving host
+# now; a routable mgmt address once the manager moves to k8s). Just another
+# *.fleet.lab like the boot/nvme/audit names. Must match the agent server cert SAN
+# (server-pki.sh AGENT_SANS).
+MGMT_HOST=manager.fleet.lab
 
 # --- NVMe/TCP target -------------------------------------------------------
 # Per-version subsystem NQN base. The long-lived target serves each rootfs

serving/docker-compose.yml

@@ -104,6 +104,9 @@ services:
       # Host-generic remote audit collector name (baked into the image config so
       # every host offloads auditd here; resolved per-environment to this host).
       - --address=/${AUDIT_HOST}/${SERVER_IP}
+      # Fleet-manager AGENT plane (:9600) — baked into the image so enrolled hosts
+      # dial the control plane by name; resolved here like the other fleet names.
+      - --address=/${MGMT_HOST}/${SERVER_IP}
       # Kiosk is LOCAL-ONLY (loads boot.fleet.lab), so sinkhole Chromium's
       # call-home to 0.0.0.0 — telemetry can't leave the lab even if the browser
       # tries. Defense-in-depth behind the in-image Chromium policy/flags; in

serving/fleet-manager/server-pki.sh

@@ -17,7 +17,7 @@
 # Usage:  server-pki.sh
 #   env overrides:
 #     OUT=./pki/agent                     output dir
-#     AGENT_SANS=diskless-server.bitasylum.org,10.0.0.1   SANs hosts dial (DNS/IP, comma-sep)
+#     AGENT_SANS=manager.fleet.lab,10.0.0.1   SANs hosts dial (DNS/IP, comma-sep)
 #     FLEET_CA_URL=https://ca.fleet.lab:8443
 #     FLEET_CA_ROOT=../dot1x-bench/pki/root_ca.crt
 #     STEP_PROVISIONER=enrol
@@ -26,7 +26,7 @@
 set -euo pipefail
 
 OUT=${OUT:-./pki/agent}
-AGENT_SANS=${AGENT_SANS:-diskless-server.bitasylum.org}
+AGENT_SANS=${AGENT_SANS:-manager.fleet.lab}
 FLEET_CA_URL=${FLEET_CA_URL:-https://ca.fleet.lab:8443}
 FLEET_CA_ROOT=${FLEET_CA_ROOT:-../dot1x-bench/pki/root_ca.crt}
 FLEET_CA_INTERMEDIATE=${FLEET_CA_INTERMEDIATE:-../dot1x-bench/pki/intermediate_ca.crt}