SYBILing nodes! 😱 Exploiting IC Network... Community Attention Required!

To try to be constructive here… it’s certainly a big concern for me as well @Lorimer. I wanted to share that I do think that there is a way to limit the risk. Steps would be:

  1. somehow agree, community-wide, which providers might be linked; possibly through an NNS vote, or maybe there is some other more efficient way
  2. we can use the this information in the DRE tooling to maximally decentralize nodes considering the potential links, as this is currently done with Provider, DC, Country, etc. We could add the potential links as a new dimension for optimizing decentralization. Some changes in the DRE tooling would be needed but I don’t think it’s too hard.

Then we could just submit proposals that reduce the number of nodes that these ā€œclustersā€ of potentially linked nodes have.
WDYT? Any suggestions how to practically implement the above? The DRE tooling changes are relatively easy, the tricky part would be mapping out links and coming to an agreement that everyone is happy with.

And additionally we need to come up with a way to identify and handle these cases in the future. We can’t expect 100% success rate since it’s not possible to prove that someone does NOT have a link to someone else. It’s only possible to prove an existing link in some cases, and have suspicion otherwise. In this thread there is a lot of suspicion and little proofs, but even suspicion is actionable IMHO. Let’s just try to be constructive and pragmatic.

14 Likes