On Wed, Jan 6, 2016 at 10:59 AM, Nawfal Abbassi Saber <nawfal.abbassi.sa...@gmail.com> wrote: > Hi FreeNet, > > I'm doing a research about peer to peer file storage systems and i would > like to know if the system FreeNet has some mechanisms to deal with churn. >
Short answer: yes, Freenet has mechanisms to deal with network churn I suppose you'd also like to learn something about those mechanism, so let me introduce you to the basic concepts. Most files on Freenet are stored with several layers of redundancy. 1. On the lowest level, for large files, Freenet uses segmentation with forward error correction to keep entire files available when some of their segments are unavailable on the network. 2. Freenet nodes store newly inserted data they consider to match their own location closely enough, so a single segment may end up stored on multiple nodes. 3. When a segment is retrieved, it will be cached by most of the nodes that relay the data to the requester. 4. After a large file is retrieved, some if its non-retrievable segments may with probability be recalculated and re-inserted by the requester, further restoring the reliability of the file. As for request routing, Freenet maintains a small-world network topology that allows for efficient routing. This topology can be maintained even under the presence of network churn. How this works depends on the particular mode of operation of Freenet, e.g. opennet vs. darknet, but it boils down to path folding and location swapping strategies. For more information, especially on the ongoing work of maintaining the network topology, please refer to the literature presented on the Freenet website [0]. Kind regards, Bert [0]: https://freenetproject.org/about.html#papers _______________________________________________ Support mailing list Support@freenetproject.org http://news.gmane.org/gmane.network.freenet.support Unsubscribe at http://emu.freenetproject.org/cgi-bin/mailman/listinfo/support Or mailto:support-requ...@freenetproject.org?subject=unsubscribe
