On November 17, 2003 08:49 pm, Martin Stone Davis wrote: > Toad wrote: > > On Mon, Nov 17, 2003 at 03:42:46PM -0800, Martin Stone Davis wrote: > >>Ken Corson wrote: > >>>Martin Stone Davis wrote: > >>>>Martin Stone Davis wrote: > >>>>>We start at the top of the list, and see who is going to make our > >>>>>will the fastest. Since our lawyer is "backed off" at the moment, we > >>>>>go with our chef. > >>> > >>>important: "at the moment" . How big do we consider this "moment" to > >>>be ? 100ms , 5 seconds, a singular point in time ? hmmm.... > >> > >>If what you're saying is that we need to include back-off time into the > >>estimate(), then I agree. Just because the node is backed off doesn't > >>mean we shouldn't consider it for routing. > > > > Eeeeeeeee. The whole point with backoff is that each request does NOT go > > to all the overloaded nodes in the RT - it's a damage limitation system, > > a la ian's socialist metaphor. > > I know. You must have misunderstood me. Let me try again: > > Recall the lawyer/chef analogy. Let's suppose that we only have one > query to make and it is the "make a will" query. Say the lawyer is > backed-off right now, but the chef is ready. However, let's say that > the lawyer is only backed off for 10 seconds, and we know that the chef > is going to take at least 11 seconds more than the lawyer to make the > will. Our best choice would then be to plan to send the "make a will" > query to the lawyer as soon as he ISN'T backed-off (i.e. in 10 seconds).
Two points. 1. we usually have several lawyers to try 2. even if a lawyer is ready he may still refuse pDNF(node) requests Which changes the picture a bit. Exactly what formula would you use? To make this work we would have to ask NGR give us a time after which the best lawyers may be ready. Ed _______________________________________________ Devl mailing list [EMAIL PROTECTED] http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/devl
