UP events solve open tickets even after bridge restarts

In-memory dedupe resets on pod restart; recovery then no-ops and leaves
stale open tickets. UP now falls back to a GLPI search for an unsolved
ticket for that monitor.
Ticket: https://projects.knownelement.com/issues/824

cmd/bridge/main.go

@@ -216,8 +216,12 @@ func (b *bridge) handle(w http.ResponseWriter, r *http.Request) {
 	case 1: // UP
 		tid, exists := b.open[k.Monitor.Name]
 		if !exists {
+			// restarts wipe the in-memory map; search GLPI for an open ticket
+			tid, _ = b.glpi.openTicketFor(k.Monitor.Name)
+			if tid == 0 {
 				return
 			}
+		}
 		delete(b.open, k.Monitor.Name)
 		go func() {
 			fu := fmt.Sprintf("RECOVERY: monitor %s back UP at %s. Ticket auto-closing as solved with recovery evidence.", k.Monitor.Name, k.Heartbeat.Time)