feat(atlas): Phase B — substrate from the live cluster

The substrate machines are now rendered from the live k3s nodes (arch, cpu, memory, GPU, k3s version), read in-cluster via the mounted ServiceAccount with a stdlib HTTP client (no client-go) and a 30s cache; falls back to the authored substrate whenever the cluster is unreachable. Placement per the homelab decision "in-cluster only for in-workload ops" — a running web app rendering live data fits. HostsFromNodes parser built test-first. RBAC (read-only node-reader SA) shipped in infra k3s/apps/cad-atlas/. Verified: go build/vet/lint(0)/test green. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-20 00:43:20 +02:00
parent a9c72d6ca8
commit 74dffc1ff3
5 changed files with 224 additions and 3 deletions
@@ -4,9 +4,12 @@ import (
 	_ "embed"
 	"encoding/json"
 	"net/http"
+	"sync"
+	"time"

 	cadatlas "git.d-ma.be/mathias/cad-atlas"
 	"git.d-ma.be/mathias/cad-atlas/internal/atlas"
+	"git.d-ma.be/mathias/cad-atlas/internal/cluster"
 )

 // atlasHTML is the Phase-A/B static shell. It fetches /api/atlas.json at load
@@ -16,7 +19,8 @@ import (
 var atlasHTML []byte

 // NewHandler serves the CAD Atlas: the shell at "/", and the sourced data at
-// "/api/atlas.json" (authored data + CI stage generated from the real cd.yml).
+// "/api/atlas.json" (authored data + CI stage generated from the real cd.yml +
+// substrate overlaid from the live cluster when reachable).
 func NewHandler() http.Handler {
 	mux := http.NewServeMux()
 	mux.HandleFunc("/", func(w http.ResponseWriter, r *http.Request) {
@@ -33,8 +37,42 @@ func NewHandler() http.Handler {
 			http.Error(w, "atlas build failed", http.StatusInternalServerError)
 			return
 		}
+		if live := liveSubstrate(); len(live) > 0 {
+			a.Substrate = live
+		}
 		w.Header().Set("Content-Type", "application/json; charset=utf-8")
 		_ = json.NewEncoder(w).Encode(a)
 	})
 	return mux
 }
+
+// substrate cache: query the cluster at most once per TTL; fall back to the
+// authored substrate (returns nil) whenever the cluster is unreachable.
+var (
+	subMu    sync.Mutex
+	subCache []atlas.Host
+	subAt    time.Time
+)
+
+const substrateTTL = 30 * time.Second
+
+func liveSubstrate() []atlas.Host {
+	subMu.Lock()
+	defer subMu.Unlock()
+	if !subAt.IsZero() && time.Since(subAt) < substrateTTL {
+		return subCache
+	}
+	subAt = time.Now()
+	raw, err := cluster.Nodes()
+	if err != nil {
+		subCache = nil
+		return nil
+	}
+	hosts, err := atlas.HostsFromNodes(raw)
+	if err != nil {
+		subCache = nil
+		return nil
+	}
+	subCache = hosts
+	return hosts
+}
@@ -114,7 +114,7 @@
 <body>
 <header>
  <h1><b>CAD</b> Atlas · From Signal to Pod</h1>
-  <span class="sub mono">one human gate · everything up- and downstream is agents · <em>v0.4 · data-driven (CI stage generated from cd.yml)</em></span>
+  <span class="sub mono">one human gate · everything up- and downstream is agents · <em>v0.5 · data-driven (CI + substrate from the live cluster)</em></span>
  <div class="controls">
    <button id="replay"><span class="dot"></span> Replay</button>
    <button id="slowmo">Slow-mo · <span id="slowState">off</span></button>
@@ -145,7 +145,7 @@
 <footer class="mono">
  CAD → CI → CD · intent→specify→dispatch · build→test→validate · deploy→ship.
  Dashed violet = feedback bus (stage 08 → TELOS: deploy outcome scored vs originating goal).
-  Data served from <code>/api/atlas.json</code> (authored <code>atlas.json</code> + CI stage generated from the live <code>cd.yml</code>).
+  Data served from <code>/api/atlas.json</code> (authored <code>atlas.json</code> + CI stage from the live <code>cd.yml</code> + substrate from the live cluster nodes).
  Phase C plugs in live reads of <code>assessor-loop</code> ledger · <code>session_log</code> · Gitea run API · Flux events.
 </footer>