A bug in TCP/Sack1 that leads to long finish time of short T
Brought to you by:
heidemann,
tom_henderson
TCP/Sack1 does not check packet to be retransmitted
when there is no new data packet to send.
This bug may leads to misunderstanding
(under-estimation) on the performance of TCP with
finite flows.
See detailed information at
http://www.cs.caltech.edu/~weixl/technical/ns2bugfixes/ns_sack1_bug_sep06/index.html
And the patch is at