Sent from my wireless phone. On Nov 19, 2012 10:24 AM, "Juiceman" <juicema...@gmail.com> wrote: > > Idea for automated seednode collection: > > Could official seednodes pass a list of second tier seednodes that newbies can try to connect to when official nodes are overloaded? > > Implementation: > > When an official seednode accepts a newbie node for announcement it checks if that node has its "be a seednode" flag set. It puts this node on a list of potential volunteers. > > After 30 minutes of not being connected to the volunteer: > The official seednode then tries to connect to the newbie to test its firewall and that it accepts an announcement attempt. If everything looks good it adds this node to a list of volunteer nodes that can be doled out to the future newbies that connect.
Clarification: the reason we wait 30 minutes is to give the new volunteers time to settle into the network and make sure they stuck around. I suggest we make it so volunteers can't dole out their own list of sub-volunteers somehow or else routing will be fubared. Perhaps when volunteer nodes get connected they fetch the latest list of official seednodes from Freenet and if they are not on it disable handing out their list of their own volunteers. Let's make it so seednodes don't hand out second tiers until they themselves have been up for 30 minutes. This gives time to check whether they are official seednodes and settle into the network. Seednodes should not persist their volunteer list past shutdown so they collect fresh volunteers and don't hand out ancient lists. > > On Nov 19, 2012 7:42 AM, "Matthew Toseland" <t...@amphibian.dyndns.org> wrote: >> >> I have merged Robert's original patch (compile fixed) for depth-first announcement onto a branch, opennet-changes. >> >> It will need to be tested thoroughly: >> - Do new nodes announce with new seeds? >> - Do new nodes announce with old seeds? >> - Do old nodes announce with new seeds? >> (Note that Update Over Mandatory relies on announcement, as well as announcing existing nodes) >> >> However, it is a very small patch, and since we wait for transfers to complete *after* we call addRefIfWanted(), it should be okay. >> >> IMHO the theoretical justification for this patch is more than adequate: >> - The nodes at the "end" of the announcement path should reply first. They are much less likely to be overloaded. >> - They are also closer to the target location, so announcement should be more effective. >> >> However, I suspect the difference in practice will be nil: >> - We will receive the same set of announcement offers, just backwards. >> - We will accept all of them because we are a newbie, and normally we get rather fewer announcement offers than our connection limit. >> >> => Getting more seednodes (preferably via an automated process) is more important. >> >> Thoughts? >> >> _______________________________________________ >> Devl mailing list >> Devl@freenetproject.org >> https://emu.freenetproject.org/cgi-bin/mailman/listinfo/devl
_______________________________________________ Devl mailing list Devl@freenetproject.org https://emu.freenetproject.org/cgi-bin/mailman/listinfo/devl
