SI
SI
discoversearch

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor.  We ask that you disable ad blocking while on Silicon Investor in the best interests of our community.  If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.
Technology Stocks : Novell (NOVL) dirt cheap, good buy?

 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext  
To: EPS who wrote (25343)2/13/1999 1:32:00 PM
From: Scott C. Lemon  Read Replies (2) of 42771
 
Hello Victor,

> February 8, 1999
>
> One Man's Dream to Spin a Faster Web
>
> By JOHN MARKOFF
>
> PALO ALTO, Calif. -- As Alan Huang sees it, the key to the
> Internet's future may lie in a simple card trick.

This is a facinating article in a couple of ways ... first, it leads me to do some research into "group theory" and it's applicability to networks. I looked into Evariste Galois and now recall the story of his duel ... now I need to find more material on his theories ...

But in looking for Terabit Routers, the company that Alan Huang founded, and I can't even find a web site. Although the competitors listed in the article seem to be alive and well.

In my opinion, the idea of improving the Internet performance issues by increasing the size of the pipes, and the "forwarding rates" of the core routers, just won't do it. To me, there are two core types of Internet data ... real-time streams ("flows"), and static objects. Today these are both made up of packets that are "flowing" through the infrastructure of the Internet.

A couple of years ago I posted on these boards my beliefs that people designing the infrastructure will one day wake up to the fact that "object routing" and real-time "flow management" will become the higher-level, more efficient ways to deal with information on the Internet. Instead of trying to make packets move faster, deal with the static objects (i.e. web pages, GIFs, JPGs, multi-media files) as whole objects ... not a bunch of packets. If the infrastructure of the Internet understands whole objects, these objects can then be leveraged for higher-level information, and also cached. And it is this ability to cache the objects that then relieves the traffic bottle necks ... it doesn't have to flow through the infrastructure of the Internet if there is a cached copy near you! ;-)

Oh yes, caching is core to this concept ... and a proxy/cache, in my opinion, *is* an "object router" .... and BorderManager and it's underlying technology are the foundation of a very high-performance, commodity hardware object router.

If object routing is implemented properly, then the only remaining "packet" traffic is real-time "flows", and other objects with a high rate of change. The "flows" are primarily your "video/voice over IP" communications or real-time broadcasts which require that the information arrives end-to-end with minimal latency. And yes, for those techies out there, even these flows could be viewed as very small objects which have a very quick expiration time ...

So cool stuff reading about these terabit routers, but I'm a fan of moving up the ISO model ...

Scott C. Lemon
Report TOU ViolationShare This Post
 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext