Transport protocols and bulk file transfer

[Yunhong Gu <>]

>
> In Section 2.3, references to SABUL and SOBAS. Could the authors of this
> protocol comment on the TCP friendliness of the protocols? Since we have the
> authors of SABUL on board, we can hear from the experts; but our experience
> with SABUL (old version) is that the protocol is not so stable and not
> designed for TCP-friendliness. But we could be wrong and also the protocol
> also could have been changed over time.
>

Dear Prof. Rhee,

The old SABUL (two years ago) was using an MIMD like control and was
designed only to be used in private networks. In addition, at that time
the implementation was one of the first several prototypes so software
issues might happen.

We have been developing new version of SABUL and its successor UDT.
(SABUL was terminated at version 2.3 and then UDT 1.0 started.) The
current UDT version is 2.0, which can be downloaded from udt.sf.net. The
new UDT code provides much better software quality.

We also have new control approaches in UDT 2.0. If you are interested, you
can find more information in this paper:
http://www.sc-conference.org/sc2004/schedule/pdfs/pap155.pdf

We have a similar definition of TCP-friendliness as you metioned in that
SIGCOMM03 paper (sigcomm 02?). Basically we want UDT to be less aggressive
than TCP in the situation when TCP works well but can be more aggressive when
TCP does NOT work well.

By the way, the reference of SABUL in the draft describes the control
mechanism used the last version of SABUL (v2.3) and this control mechanism
havs been improved in UDT 2.0, which is described by the above SC04 paper.

Thanks,
Gu