app/webui: fix component-status.json multi-process clobber and unbounded growth

The long-lived bee-web process (writing PSU/kmsg watchdog records ~every
60s) and each short-lived "bee bee-worker" SAT-task subprocess each held
an independent in-memory copy of component-status.json. Whichever saved
last won outright, silently erasing whatever the other had just written —
e.g. a GPU SAT task's pcie:gpu:nvidia result vanishing the next time the
PSU watchdog ticked. ComponentStatusDB.Record now reloads on-disk state
(keyed by newer LastCheckedAt) before merging its own update.

Also stops re-logging identical repeat observations to History: the
ingest contract defines status_history as a transition log ("История
переходов статусов"), not a per-poll journal, but Record appended one
entry per call regardless — a continuously-polled PSU grew an unbounded
run of identical "still OK" entries. Now only appends when a source's
last recorded status for a key actually changes.

The PSU watchdog itself now backs off (60s -> doubling, capped at 30min)
while steady and resets to 60s the moment any PSU's status changes, cutting
ipmitool shellouts and file writes for a fleet that's been stable for a
while.

Also adds the missing "raid" case to the component-detail API (was
returning 404 for any RAID card's detail click on /topo).

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

audit/internal/webui/api.go

@@ -1851,6 +1851,9 @@ func (h *handler) handleAPIComponentDetail(w http.ResponseWriter, r *http.Reques
 	case "psu":
 		title = "PSU"
 		prefixes = []string{"psu:"}
+	case "raid":
+		title = "RAID"
+		prefixes = []string{"pcie:raid:"}
 	default:
 		http.NotFound(w, r)
 		return

audit/internal/webui/health_poller.go

@@ -11,17 +11,32 @@ import (
 	"bee/audit/internal/collector"
 )
 
-const healthPollInterval = 60 * time.Second
-const psuIPMITimeout = 15 * time.Second
+const (
+	healthPollIntervalMin = 60 * time.Second
+	// healthPollIntervalMax caps the backoff below: a PSU that has been
+	// steady for a while is polled at most this rarely, so a real failure
+	// is still caught within a bounded window even after a long quiet
+	// stretch.
+	healthPollIntervalMax = 30 * time.Minute
+	psuIPMITimeout        = 15 * time.Second
+)
 
 // healthPoller runs periodic health checks for hardware components that do not
 // emit kernel log events (e.g. PSU). Results are written to ComponentStatusDB.
+//
+// The poll interval backs off (doubling, capped at healthPollIntervalMax)
+// each tick where no PSU's status changed, and resets to
+// healthPollIntervalMin the moment any PSU does change — a steady-state PSU
+// bank doesn't need re-checking every 60s forever (each poll is a
+// component-status.json write and an ipmitool shellout), while a PSU that
+// just started flapping gets caught quickly again.
 type healthPoller struct {
 	statusDB *app.ComponentStatusDB
+	interval time.Duration
 }
 
 func newHealthPoller(statusDB *app.ComponentStatusDB) *healthPoller {
-	return &healthPoller{statusDB: statusDB}
+	return &healthPoller{statusDB: statusDB, interval: healthPollIntervalMin}
 }
 
 func (p *healthPoller) start() {
@@ -29,16 +44,34 @@ func (p *healthPoller) start() {
 }
 
 func (p *healthPoller) run() {
-	ticker := time.NewTicker(healthPollInterval)
-	defer ticker.Stop()
-	for range ticker.C {
-		p.pollPSU()
+	timer := time.NewTimer(p.interval)
+	defer timer.Stop()
+	for range timer.C {
+		p.interval = nextHealthPollInterval(p.interval, p.pollPSU())
+		timer.Reset(p.interval)
 	}
 }
 
-func (p *healthPoller) pollPSU() {
+// nextHealthPollInterval computes the next poll interval given whether the
+// last poll observed any status change: reset to the fast floor on change,
+// otherwise double (capped) toward the slow ceiling.
+func nextHealthPollInterval(current time.Duration, changed bool) time.Duration {
+	if changed {
+		return healthPollIntervalMin
+	}
+	next := current * 2
+	if next > healthPollIntervalMax {
+		next = healthPollIntervalMax
+	}
+	return next
+}
+
+// pollPSU polls PSU status via ipmitool and records it to statusDB. Returns
+// true if any PSU's status differs from what statusDB currently has for it,
+// which run uses to reset the backoff.
+func (p *healthPoller) pollPSU() bool {
 	if p.statusDB == nil {
-		return
+		return false
 	}
 	ctx, cancel := context.WithTimeout(context.Background(), psuIPMITimeout)
 	defer cancel()
@@ -49,21 +82,25 @@ func (p *healthPoller) pollPSU() {
 	if err := cmd.Run(); err != nil {
 		// IPMI not available or not a server — skip silently.
 		slog.Debug("health poller: ipmitool sdr unavailable", "err", err)
-		return
+		return false
 	}
 
 	slots := collector.PSUSlotsFromSDR(out.String())
 	if len(slots) == 0 {
-		return
+		return false
 	}
 
 	const source = "watchdog:psu"
+	changed := false
 	for slot, psu := range slots {
 		key := "psu:" + slot
 		status := psu.Status
 		if status == "" {
 			status = "Unknown"
 		}
+		if prev, ok := p.statusDB.Get(key); !ok || prev.Status != status {
+			changed = true
+		}
 		detail := ""
 		switch status {
 		case "Critical":
@@ -73,4 +110,5 @@ func (p *healthPoller) pollPSU() {
 		}
 		p.statusDB.Record(key, source, status, detail)
 	}
+	return changed
 }

audit/internal/webui/health_poller_test.go

@@ -0,0 +1,54 @@
+package webui
+
+import (
+	"path/filepath"
+	"testing"
+	"time"
+
+	"bee/audit/internal/app"
+)
+
+func TestNextHealthPollIntervalBacksOffAndResetsOnChange(t *testing.T) {
+	interval := healthPollIntervalMin
+	for i := 0; i < 10; i++ {
+		interval = nextHealthPollInterval(interval, false)
+		if interval > healthPollIntervalMax {
+			t.Fatalf("interval=%s exceeded cap %s after %d steady ticks", interval, healthPollIntervalMax, i+1)
+		}
+	}
+	if interval != healthPollIntervalMax {
+		t.Fatalf("interval=%s want cap %s after repeated steady ticks", interval, healthPollIntervalMax)
+	}
+
+	// A change resets straight back to the fast floor, regardless of how
+	// far the backoff had climbed.
+	if got := nextHealthPollInterval(interval, true); got != healthPollIntervalMin {
+		t.Fatalf("interval after change=%s want floor %s", got, healthPollIntervalMin)
+	}
+}
+
+func TestHealthPollerPollPSUUnavailableReturnsFalseWithoutPanic(t *testing.T) {
+	// ipmitool is not present in this test environment, so pollPSU must
+	// degrade to a no-op (false, no crash) rather than erroring out the
+	// poller loop.
+	db, err := app.OpenComponentStatusDB(filepath.Join(t.TempDir(), "component-status.json"))
+	if err != nil {
+		t.Fatal(err)
+	}
+	p := newHealthPoller(db)
+	if p.interval != healthPollIntervalMin {
+		t.Fatalf("initial interval=%s want floor %s", p.interval, healthPollIntervalMin)
+	}
+	if got := p.pollPSU(); got {
+		t.Fatalf("pollPSU()=%v want false when ipmitool is unavailable", got)
+	}
+}
+
+func TestHealthPollIntervalBoundsAreSane(t *testing.T) {
+	if healthPollIntervalMin >= healthPollIntervalMax {
+		t.Fatalf("min %s must be less than max %s", healthPollIntervalMin, healthPollIntervalMax)
+	}
+	if healthPollIntervalMin != 60*time.Second {
+		t.Fatalf("min=%s want 60s (unchanged default poll rate)", healthPollIntervalMin)
+	}
+}