[e2e] Net-unfriendly solutions (was Is a non-TCP solution dead?)

[Spencer Dawkins]

I'm not the hall monitor for TCP implementations, or for PEPs,
but I would add that I don't know how you put together an
architecture like this without TCP ack-spoofing. An unmodified
TCP sender isn't going to add packets to the network without
acks saying "packets have left the network", no matter how
cool(?) the protocol between PEPs might be.

And, as long as support for application-level command pipelining
is limited in widely-deployed browsers (for instance),
application-level ack-spoofing may be required as well.

There are applications with well-defined cache-interaction
semantics, but (for instance) I don't know how you tell SMTP
that the e-mail it just sent, and received acks for, was lost by
an ack-spoofing PEP, in any way that will result in a
non-user-initiated retransmission.

So, one wonders about NetLightning's effectiveness for anything
other than HTTP (are there other applications?)...

And NetLightning is not the first one I've seen (not to name
names)...

Spencer

--- "Armando L. Caro Jr." <acaro at mail.eecis.udel.edu> wrote:
> What worries me is how much praise such violations receive.
> Network World
> (issue 04/21/03) has an article that praises a startup
> called Netli, which has developed a producted called
> NetLightning.
> NetLightning is a CDN architecture which essentially
> establishes three
> transport layer connections: (1) client to local VDC (Virtual
> Data
> Center), (2) local VDC to remote AAP (Application Access
> Point), and (3)
> remote AAP to HTTP server. Connections (1) and (3) use the
> normal
> HTTP/TCP, but connection (3) uses Netli's prioprietary
> transport protocol.
> 
> According to Netli, their transport protocol "uses the
> reliable, secure,
> and network-friendly techniques of TCP, but is enhanced to
> provide optimal
> performance for communication between a VDC in one geographic
> region and
> an AAP in another." However, Tim Greene of Network World
> reports more
> details which reveal just how network-friendly Netli's
> protocol is. For
> example, Greene reports "that the [Netli] protocol can reduce
> from 31 to
> two the number of round-trip interactions needed to send a 70K
> byte Web
> page with 25 objects on it." Also, the Netli Protocol
> "eliminates the
> slow-start feature of TCP," because "these and other
> refinements of TCP
> reduce the time it takes to download pages and to interact
> with Web
> applications."
> 
> NetLightning could become a problem if it becomes deployed as
> widely as
> Alkami's CDN technology. With customers such as HP and
> investors such as
> Nokia, it is likely that NetLightning will spread and Netli's
> competitors
> will adopt similar techniques.
> 
> FYI: Netli are present at the Networld+Interop in Las Vegas,
> and they have
> a marketing talk today at 2pm PST.
> 
> Network World Article:
> http://www.nwfusion.com/news/2003/0421netli.html
> 
> Netli's Website:
> www.netli.com
> 
> ~armando