-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA512 Natanael wrote: > Den 16 feb 2015 12:53 skrev "Adonay Sanz" <adonay.s...@gmail.com>: >> >> Hi Natanael Thanks for answering. Really I'm noob in this area. >> >> I wanted to use I2P version. So, it's no possible way present or >> future > to block it? It's a really threat? >> Can you explainme better what it means "configured to mimic >> Freenet"? > > Freenet caches files you request, and caches some files other > people request via you. Then it replies to requests from others > and forwards requests as well. Tahoe-LAFS in I2P mode mimics this > behaviour.
No it doesn't. Data storage and caching for Tahoe-LAFS in I2P is identical to standard Tahoe-LAFS. > > In I2P it isn't certain you'll be able to find honest nodes which > have the data you are asking for. They can send fake replies, and > even though your software will know the replies are wrong it will > have to spend a lot of effort on searching for the correct data. > So a Sybil attack can be used for censorship. > > But if you know of honest nodes, and the honest nodes are well > connected in the network, then the requests will reach those nodes > and the correct reply will come back. But the fake nodes can > overload the honest nodes to slow things down. I don't see how this is any different to standard Tahoe-LAFS. The only difference between standard and I2P Tahoe-LAFS is the addition of HTTP proxy support (to enable I2P communication, although better ways are in discussion in Tahoe-LAFS and Foolscap Trac tickets), and support for multiple introducers (so that one going down does not kill the entire I2P Tahoe-LAFS network). Other than that, there is no difference. Non-honest Tahoe-LAFS nodes are perfectly capable of joining a standard Tahoe-LAFS network and sending fake replies. So any Sybil attack on the I2P Tahoe-LAFS network would be directly applicable to a standard Tahoe-LAFS network. I should point out that there is a distinct difference between a Sybil attack on Tahoe-LAFS, and a Sybil attack on I2P itself. An adversary _could_ execute a Sybil attack on I2P, but the most they could do to affect the I2P Tahoe-LAFS network is attempt an Eclipse attack on the I2P Destinations of Tahoe-LAFS nodes. The I2P Tahoe-LAFS network uses B32 addresses to refer to nodes, and an Eclipse attack could cause lookups of these B32s to fail. The scale needed to block out all regular nodes would be completely impractical, however, given that the I2P Tahoe-LAFS network is one of the largest known Tahoe-LAFS networks, and an attacker would need to conduct a partial-keyspace attack against each legitimate node independently. They could instead target the introducers, but that would be noticed as an attack _very_ quickly. And even this attack has an equivalent for standard Tahoe-LAFS networks: an adversary could cause the DNS lookups of hostnames for regular Tahoe-LAFS nodes to fail, or (at a lower level) prevent connections to known Tahoe-LAFS node IPs. So I still don't see how I2P Tahoe-LAFS is any weaker that standard Tahoe-LAFS. str4d > > > > _______________________________________________ tahoe-dev mailing > list tahoe-dev@tahoe-lafs.org > https://tahoe-lafs.org/cgi-bin/mailman/listinfo/tahoe-dev > -----BEGIN PGP SIGNATURE----- iQIcBAEBCgAGBQJU4nJMAAoJEIA97kkaNHPnKjsP/3hSgwufcHiXVIWld97Lf/Ad 2GDI4P/XMteEz4C71hEgXF1bQRrG66HPuBjph8POjK0x3w0i+ajZ/EpV9sNpapIu C0iJ/3V0Q+VXDVxb0yqZQ2wvhF903H8hRgK09Cx7oFBjpf8PjxUE11j4h9RCEBxs SHCJDcI39uJqbSPDWvFXDbEs83hc9+IA+GsIUnXXdPKXUAYi/CR45tnxj0524FUi Scy598JlaQ7JuvkXT4gZkDvv+/3KwHt1TdBZTMHZKqo8Jy89QTudQZBQ8b7oEfz4 ZZm0tWUjnnt+cM6jJhmBMc2jBQvnSjMOSojnP3DRfP0MmQt3ewdKKIp3qmmYPsq4 wjPJqbWuvWz1v1sYZ9REzB+LO9e6gv+iJ8IQnj1XaN1T6ThfUThMUTAmi9OC3WIi mkxQ1jMAbEs6XJ+hGBKYQ8wHa/r8iI2KhTj/jI1QH9aFCyj1i3pkHVs5ZwIdHZK/ F9tyCrDYhlZPtRzmnt9oTOdv4njKPbgbCgUWgS3rxIh5eCLB3QTZUDw39Ldh7A7l wjjfe+LfTFFail6Ncn+m2hzpupnZsq9KZy5wxRbc8HxshZl0zvcXcl1emwcPLmOK CvZn0L6D7wM2PQi2epBlBIxR0pSLYYRga07g/DdmtXUfZ455fH1v9MQ7kxicdQMN 4H5qpP+35oIAfhZfcoaF =rhmK -----END PGP SIGNATURE----- _______________________________________________ tahoe-dev mailing list tahoe-dev@tahoe-lafs.org https://tahoe-lafs.org/cgi-bin/mailman/listinfo/tahoe-dev
