refactor(knowledge): migrate deleted-item cleanup from polling scheduler to event-driven worker

Replaces the scheduler-polled KnowledgeDeletedCleanupWorker with an
event-driven KnowledgeDeletedCleanupEventWorker in tenant-api; cleanup
events are now enqueued transactionally at delete time instead of swept
periodically. Also fixes statement_timeout parameterization in
MonitoringRetentionWorker and propagates errors from cleanupDeletedKnowledgeItem.

Co-Authored-By: Claude Sonnet 4.6 (1M context) <noreply@anthropic.com>

server/internal/scheduler/knowledge_deleted_cleanup_worker.go

@@ -1,58 +0,0 @@
-package scheduler
-
-import (
-	"context"
-	"time"
-
-	tenantapp "github.com/geo-platform/tenant-api/internal/tenant/app"
-)
-
-type KnowledgeDeletedCleanupWorker struct {
-	service  *tenantapp.KnowledgeService
-	interval time.Duration
-}
-
-func NewKnowledgeDeletedCleanupWorker(service *tenantapp.KnowledgeService) *KnowledgeDeletedCleanupWorker {
-	return &KnowledgeDeletedCleanupWorker{
-		service:  service,
-		interval: 15 * time.Second,
-	}
-}
-
-func (w *KnowledgeDeletedCleanupWorker) Start(ctx context.Context) {
-	go w.Run(ctx)
-}
-
-func (w *KnowledgeDeletedCleanupWorker) Run(ctx context.Context) {
-	if w == nil || w.service == nil {
-		return
-	}
-	w.run(ctx)
-}
-
-func (w *KnowledgeDeletedCleanupWorker) run(ctx context.Context) {
-	w.service.RunDeletedCleanupSweep()
-
-	ticker := time.NewTicker(w.interval)
-	defer ticker.Stop()
-
-	for {
-		select {
-		case <-ctx.Done():
-			return
-		case <-ticker.C:
-			w.service.RunDeletedCleanupSweep()
-		}
-	}
-}
-
-func (w *KnowledgeDeletedCleanupWorker) RunOnce(ctx context.Context) (map[string]any, error) {
-	if w == nil || w.service == nil {
-		return map[string]any{"skipped": true, "reason": "worker_not_ready"}, nil
-	}
-	startedAt := time.Now()
-	w.service.RunDeletedCleanupSweep()
-	return map[string]any{
-		"duration_ms": time.Since(startedAt).Milliseconds(),
-	}, nil
-}

server/internal/scheduler/monitoring_retention_worker.go

@@ -70,7 +70,7 @@ func (w *MonitoringRetentionWorker) RunOnce(ctx context.Context, run JobRunConte
 	defer conn.Release()
 
 	if statementTimeout > 0 {
-		if _, err := conn.Exec(ctx, `SET statement_timeout = $1`, int(statementTimeout.Milliseconds())); err != nil {
+		if _, err := conn.Exec(ctx, `SELECT set_config('statement_timeout', $1, false)`, fmt.Sprintf("%dms", statementTimeout.Milliseconds())); err != nil {
 			return stats, fmt.Errorf("set statement timeout: %w", err)
 		}
 		defer conn.Exec(context.Background(), `RESET statement_timeout`)