Dan Wing <dw...@vmware.com> writes: > The paper "On the Impact of Packet Spraying in Data Center Networks" > (*) shows impact of packet spraying with TCP and seems to show good > results without the sequence number proposed in the Internet Draft. > Can the authors comment on that paper or on the need for the sequence > number and group flow identifier, proposed in your I-D?
I've heard people mention that if TCP uses timestamps, the effect of packet reordering can be largely eliminated. Is this actually true, and if so, does anyone have a reference? Dale _______________________________________________ nvo3 mailing list nvo3@ietf.org https://www.ietf.org/mailman/listinfo/nvo3