fix(scheduler): resolve SetSnapshot race conditions and remove usage throttle

Backend: Fix three race conditions in SetSnapshot that caused account
scheduling anomalies and broken sticky sessions:
- Use Lua CAS script for atomic version activation, preventing version
  rollback when concurrent goroutines write snapshots simultaneously
- Add UnlockBucket to release rebuild lock immediately after completion
  instead of waiting 30s TTL expiry
- Replace immediate DEL of old snapshots with 60s EXPIRE grace period,
  preventing readers from hitting empty ZRANGE during version switches

Frontend: Remove serial queue throttle (1-2s delay per request) from
usage loading since backend now uses passive sampling. All usage
requests execute immediately in parallel.

backend/internal/service/scheduler_cache.go

@@ -59,6 +59,8 @@ type SchedulerCache interface {
 	UpdateLastUsed(ctx context.Context, updates map[int64]time.Time) error
 	// TryLockBucket 尝试获取分桶重建锁。
 	TryLockBucket(ctx context.Context, bucket SchedulerBucket, ttl time.Duration) (bool, error)
+	// UnlockBucket 释放分桶重建锁。
+	UnlockBucket(ctx context.Context, bucket SchedulerBucket) error
 	// ListBuckets 返回已注册的分桶集合。
 	ListBuckets(ctx context.Context) ([]SchedulerBucket, error)
 	// GetOutboxWatermark 读取 outbox 水位。

backend/internal/service/scheduler_snapshot_hydration_test.go

@@ -44,6 +44,10 @@ func (c *snapshotHydrationCache) TryLockBucket(ctx context.Context, bucket Sched
 	return true, nil
 }
 
+func (c *snapshotHydrationCache) UnlockBucket(ctx context.Context, bucket SchedulerBucket) error {
+	return nil
+}
+
 func (c *snapshotHydrationCache) ListBuckets(ctx context.Context) ([]SchedulerBucket, error) {
 	return nil, nil
 }

backend/internal/service/scheduler_snapshot_service.go

@@ -544,6 +544,9 @@ func (s *SchedulerSnapshotService) rebuildBucket(ctx context.Context, bucket Sch
 	if !ok {
 		return nil
 	}
+	defer func() {
+		_ = s.cache.UnlockBucket(ctx, bucket)
+	}()
 
 	rebuildCtx, cancel := context.WithTimeout(ctx, 30*time.Second)
 	defer cancel()