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 : The *NEW* Frank Coluccio Technology Forum
 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext  
To: Frank A. Coluccio who wrote (174)6/11/2000 2:48:00 PM
From: ftth  Read Replies (2) of 46821
 
OK, but I still don't get that from reading that person's post. They seem to be claiming that some massively parallel, *distributed* (across a backbone) architecture is the way, and this seems nearly an opposite treatment from processing classified/labelled streams. Massively paralleled distributed routing across the internet defies logic, at least the way I interpret that phrase. There'd be so much overhead with "state" information flying back and forth and sideways among routers as they try and schedule the distribution of the computational load amongst themselves. This is not the same problem as some massively paralleled distributed computation across a WAN of some "grand challenge" problem like breaking a crypto algorithm. It's also not the same problem as *local* massively parallel computation, which is a more realistic way to handle the problem.

re: >>One of his points was that you simply don't. You instead treat entire flows in aggregate, comprised of groups <which perhaps have been labeled> of like services with similar needs, as such, but not at the granularity of an individual user's packet.

That's essentially what I'm advocating with the lambda classes. Any router outside the local AS doesn't need to know or care what's inside the classified stream; it just needs to know what class it derived from. Merely making that realization (about the value of classified flow aggregation) doesn't really reduce the complexity of the problem, though. It just shifts it into a realm where it is more solvable. It's still a complicated problem. Would you agree?
Report TOU ViolationShare This Post
 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext  

HomeMailHotSubjectMarksPeopleMarksKeepersSettings
Terms Of UseContact UsCopyright/IP PolicyPrivacy PolicyAbout UsFAQAdvertise on SI
© 2025 Knight Sac Media.  Data provided by Twelve Data, Alpha Vantage, and CityFALCON News