I have wondered about a specific kind of attack against freenet, and if this would
even be
effective or trivial to protect against.
Lets say a custom client is written that allows one to connect to multiple freenet
nodes
in various locations. A person uses this node to upload some data at specific points,
lets say a number of 10M files.
This client then starts requesting this bogus data at other locations from other nodes,
with the intent of having intervening nodes cache this data in their data stores.
This data would appear popular, and would propagate through caches, and if the caches
were
full, would force existing data that did not appear popular out of the cache.
Over time, lets say a few days, and 500 of these 10M files later, it would appear to me
that a well written and connected client of this type could force a large amount of
content out of Freenet and replace it with this bogus data.
How quickly and effectively this could be done depends on the cache sizes of each node,
how connected the client is, and how much bandwidth is available. All in all though it
seems to me that this would be possible, perhaps even quickly and easily.
Now, regarding protections against this kind of attack, increasing data stores to a
given
percentage of free disk space would certainly help. Some nodes may have 20G of cache.
But
this still propagates a large amount of crap through the network.
Any thoughts on this?
Best regards...
_______________________________________________
Freenet-dev mailing list
[EMAIL PROTECTED]
http://lists.sourceforge.net/lists/listinfo/freenet-dev
