feat(core): serve status as JSON on the observability server

core/observability.go

@@ -44,6 +44,7 @@ func (n *Nylon) startObservability() error {
 	mux.HandleFunc("/readyz", n.handleReady)
 	mux.HandleFunc("/metrics", n.handleMetrics)
 	mux.HandleFunc("/discovery", n.handleDiscovery)
+	mux.HandleFunc("/status", n.handleStatusJSON)
 
 	obs := &observabilityServer{
 		listener: listener,
@@ -131,6 +132,23 @@ func (n *Nylon) handleMetrics(w http.ResponseWriter, r *http.Request) {
 	writePrometheusMetrics(w, status)
 }
 
+func (n *Nylon) handleStatusJSON(w http.ResponseWriter, r *http.Request) {
+	ctx, cancel := context.WithTimeout(r.Context(), observabilityTimeout)
+	defer cancel()
+	status, err := n.statusSnapshot(ctx)
+	if err != nil {
+		http.Error(w, "status unavailable", http.StatusServiceUnavailable)
+		return
+	}
+	data, err := pjMarshal.Marshal(status)
+	if err != nil {
+		http.Error(w, err.Error(), http.StatusInternalServerError)
+		return
+	}
+	w.Header().Set("Content-Type", "application/json")
+	_, _ = w.Write(data)
+}
+
 func (n *Nylon) handleDiscovery(w http.ResponseWriter, _ *http.Request) {
 	_, port, err := net.SplitHostPort(n.LocalCfg.ObservabilityAddr)
 	if err != nil {

core/observability_test.go

@@ -84,3 +84,14 @@ func TestPrometheusMetrics(t *testing.T) {
 	require.NotContains(t, output, `nylon_wireguard_peer_latest_handshake_seconds{peer="bob"} 1.786`)
 	require.Equal(t, 1, strings.Count(output, "# HELP nylon_wireguard_peer_transmit_bytes_total "))
 }
+
+func TestObservabilityStatusUnavailable(t *testing.T) {
+	ctx, cancel := context.WithCancelCause(context.Background())
+	defer cancel(nil)
+	// nothing runs the dispatch loop, so the snapshot times out
+	n := &Nylon{Context: ctx, DispatchChannel: make(chan func() error)}
+
+	rec := httptest.NewRecorder()
+	n.handleStatusJSON(rec, httptest.NewRequest(http.MethodGet, "/status", nil))
+	require.Equal(t, http.StatusServiceUnavailable, rec.Code)
+}

docs/guides/observability.mdx

@@ -21,6 +21,7 @@ Restart nylon after changing the node configuration. You can then check the loca
 curl http://127.0.0.1:9090/healthz
 curl http://127.0.0.1:9090/readyz
 curl http://127.0.0.1:9090/metrics
+curl http://127.0.0.1:9090/status
 ```
 
 :::caution
@@ -34,6 +35,7 @@ The observability server does not provide authentication or TLS. Bind it to a tr
 | `/healthz` | Returns `200 OK` while the daemon is running and `503 Service Unavailable` while it is shutting down. |
 | `/readyz` | Returns `200 OK` when nylon's internal event loop responds, or `503 Service Unavailable` if it cannot respond within one second. It does not check whether every peer is reachable. |
 | `/metrics` | Returns metrics in the Prometheus text exposition format. Returns `503 Service Unavailable` if nylon cannot collect a status snapshot within one second. |
+| `/status` | Returns the node's status as JSON, the same object `nylon status --json` prints under `status`. Returns `503 Service Unavailable` if nylon cannot collect a snapshot within one second. |
 | `/discovery` | Returns Prometheus HTTP service-discovery targets for the addresses in `central.yaml`. |
 
 ## Configure Prometheus

docs/reference/config.mdx

@@ -22,7 +22,7 @@ no_net_configure: false # if true, nylon won't touch system routes or interfaces
 log_path: "" # write logs to this file (empty = stderr only)
 interface_name: "" # override the interface name (default: "nylon", or utunX on macOS)
 dns_resolvers: [] # DNS servers for nylon's own lookups, e.g. ["1.1.1.1:53"]
-observability_addr: "" # e.g. "0.0.0.0:9090"; enables /metrics, /healthz, /readyz, and /discovery
+observability_addr: "" # e.g. "0.0.0.0:9090"; enables /metrics, /healthz, /readyz, /discovery, and /status
 
 # Bootstrap: fetch central.yaml from a remote bundle on first start
 dist: