I'm familiar with Menstat from a distance.
I was intrigued by the promise of the XTP protocol,
and encourage Goddard Space Flight Center to
investigate/test this protocol (which we finally did).
The results looked very promising; however, the tester had
reservations about the quality of the TCP/IP and TCP/IP SACK
stacks.
I do not know of any testing at GSFC with SkyX.
The results at the URL look like the results
generated using XTP (which if the case is only moderately
deceiving).
The problem with XTP was that it was not available
on many platforms. It is also questionable how
it will react when running with (against?) TCP/IP
streams.
It looks like Menstat has come up with a viable
idea by employing XTP over the satellite link and
providing a gateway to interoperate in the TCP world.
If it works for you, then more power to yah.
Sincerely
Clayton Sigman
-----Original Message-----
From: Uplinx <[email protected]>
To: [email protected] <[email protected]>
Date: Wednesday, December 16, 1998 4:12 PM
Subject: Re: TCP over IP over PPP (over Satellite)
>Dear TCPSAT members-
>
>I am involved in a startup that is a reseller of HNS satellite products.
We
>currently resell the Hughes Network Systems PES 5000 (with 512/128
segments
>from Galaxy VII). We provide Internet access and WAN's for the education
>market. In the first quarter of 99, we plan to switch to DirecPC
Enterprise
>Edition, which has been billed to deliver up to 24M/256k. I have been
>watching discussions on this newsgroup from afar, as I have not yet had
much
>to contribute (I am less a technician and more of a sales/marketer). Up
>until two weeks ago, our performance was terrible, slower than a bad
dial-up
>line. I had solicited HNS to provide us with solutions that would change
>the behavior of TCP over satellite, with no tangible results until
recently.
>I first came to them with a product called Mentat SkyX
>(http://www.mentat.com/skyx/skyx.html) that I had found. Mentat wants to
>sell the source code, but does not yet have a software package. Upon this
>they countered with a software solution called Fourelle Venturi
>(http://www.fourelle.com/venturi.html), which our company put into testing
>last week (HNS decided to endorse this product in Nov.). The software, in a
>simple description, converts TCP/IP into UDP at the server prior to travel
>over satellite. It then is converted back to TCP/IP by software at the
>end-user point. There are other things the software does in regards to
>error correction to counter the indifference of UDP as I understand it. The
>results were absolutely amazing. We installed the software at a Tennessee
>school that also currently has a T1 feed to the Internet. Side by side,our
>512k satellite feed consistently beat the T1-fed machines downloading web
>pages. While I suspect there are faster T1 environs out there (the school
>is on the TN board of Ed network, which is probably quite conjested, and it
>is filtered), the difference in performance is like night and day.
>
>I would like to know if anyone has experimented with Venturi, Mentat, or
any
>other similar applications, and if so, what results they had. I think I
>may have found one of the best available, but I could be wrong.
>Additionally, I would be happy to give anyone that wants it contact info
for
>Venturi. If you contact them through their website, please mention that I
>referred you (I am always looking for a way to negotiate a price break). I
>have had no conversations with Venturi, but I will soon. I cannot relate
>how good they are with support, other issues, or even pricing, yet. The
>software was given to me as a test by HNS (it expires in a month).
>
>Thank you, and best of luck to all.
>
>Jeremy McDevitt
>
>Uplinx Technologies
>[email protected]
>
Clayton Sigman
ESDIS Systems Engineer for Mission Networks
NASA GSFC Code 585
Mailstop Code 423
Greenbelt, MD 20771
Ph: (301) 614 5309 Fax: (301) 614 5267
email: [email protected]
Motto: The best plans can't be a little bit of dumb luck!
This archive was generated by hypermail 2b29 : Mon Feb 14 2000 - 16:14:52 EST