Re: [PD] Performance with multiple paths for PD