On Wed, Feb 21, 2001 at 06:58:41PM +0000, Theodore Hong wrote: > > Now as for a fix. I think we should just set a limit on the number of > > references any individual node can get to a percentage of the total > > number of references. On reference add if the number of references to > > the node > total references / x fail. x should be the number of other > > nodes we want to have references too. > > I don't think this is really necessary -- my guess is that the reason you > only see references to this one node is that it is the only one which is > consistently alive and returning data, whereas all the other ones keep > going down and getting pruned. In which case artificially preserving > references to those other nodes won't help you any. Hopefully this > behavior will change when nodes become more reliable.
I still disagree somewhat... These relatively massives nodes with 2gb datastores probably have copies of a good percentage of *all* the data on Freenet. Any time any single node can successfully process a good portion of the requests it gets on it's own you are going to screw up the references to it. However now that I think about it this is only a problem while any single node can mirror a good percentage of the total freenet data out there. When this isn't true we won't have this problem. -- retep at penguinpowered.com http://retep.tripod.com -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 232 bytes Desc: not available URL: <https://emu.freenetproject.org/pipermail/devl/attachments/20010221/11e2d63d/attachment.pgp>
