[e2e] TCP out of order

[Craig Partridge]

In message <LAW2-F943ZHtMkcVG8v00009e25 at hotmail.com>, "kalyan uppalapati" write
s:

>1) why TCP behaves so badly, when we use multipath network with a small 
>delay difference in one link ??
>the throghput drops gradually.

How large is the delay difference measured in bits?  Is it big enough to
cause ack reordering (as the message subject implies) or even packet
reordering?

If you get packet reordering, you'll get duplicate acks, which can cause
spurious fast retransmits and have TCP unnecessarily reduce its window
size.

If you get ack reordering, TCP will be more bursty, which can cause
greater packet loss (again, window size suffers) and/or packet reordering.

Are either of these syndromes consistent with what you see?  Do you have
packet traces of the interactions you're trying to puzzle out?

Thanks!

Craig