This is a read-only archive. Find the latest Linux articles, documentation, and answers at the new!

not dead yet

Posted by: Anonymous [ip:] on November 11, 2007 01:03 AM
Now and then I've heard of systems being over loded with trafic making them slow or unresponsive. I'm not that familiar with internet nodes yet, either the node or IP end point should have enough persistance or cache ageing to identify a most requested site. A variation would be to have a server create a load hierarchy that identifies the highest trafic source. Once the high trafic node is identified, it is turned into a proxy server or provides a stock ticker type broadcasting application that responds with only one request and the number of such requests from the node. Or embarks upon update mode which responds with number of requests after each update. All requests are reflected back with the currently updated state. Thus the load moves toward its source distributing demand on the ultamate source. In a simmilar way different media such as HD radio, satilite, ham, a puplic service TV radio station could provide either free or to the highest bid broadcast load dampening. All these media would have cross platform formates that allowed each of them to broadcast content from the other. A satilite service mite at a given channel broadcast the top requested web content from a given search engine. It would be formated for reception and interface with a pc or media TV application for brousing allbeit one directional- there would be no choices it would just be a headline service. Never the less the channel could be perchased temporarily to broadcast sporting events or the most requested item on the internet allowing a person to change the content source. They would get the same service yet help reduce the request load by using the redirected source. There could be a feed back feature that the end user could use to send a reciept conformation that didn't require a response but help the data source track use or determine cut off points for useing the feature based upon the amount of use.


Return to CellWriter: Open source handwriting recognition for Linux